Does artificial aging correctly predict the long-term in-vivo degradation behavior in zirconia hip prostheses?

Abstract

BACKGROUND:

Accelerated hydrothermal aging has long been one of the most widely accepted quality control tests for simulating low-temperature degradation (LTD) in zirconia-containing implants used in total hip arthroplasty (THA). However, it is still unclear how much consistency there is between the experimental prediction from the internationally-standardized tests and the actual measurements from surgically-removed implants after a long period of implantation. This question is fundamentally related to a lack of understanding of mechanical/tribological contribution to the in-vivo LTD kinetics.

OBJECTIVE:

The main purpose of this study is to validate the clinical relevance of standardized accelerated aging by comparing artificially-aged and in-vivo used prostheses, and to clarify the long-term effects of in-vivo mechanics/tribology on the LTD progression upon service in the body environment.

METHODS:

Surface magnitudes of phase transformation and residual stress in zirconia femoral head retrievals (13.1–18.4 yrs) were evaluated by using confocal Raman microspectroscopy.

RESULTS:

The long-term aging behavior in unworn head surface was in agreement with the experimental prediction estimated as 1 h aging at 134 °C = 4 years in-vivo. However, the current aging protocols based on ASTM and ISO criteria were not accurately predictive for the worn surfaces, and the tribologically-induced phase transformation and tensile stress were up to 6.5-times and 3.3-times higher than the environmentally-induced ones.

CONCLUSION:

Our study suggests that wear/scratching, frictional heating, tribochemical reactions, and metal transfer may become far more intense triggers to phase transformation than the mere exposure to body fluid.

Keywords

Aging kinetics zirconia femoral head phase transformation residual stresses

1. Introduction

Yttria-stabilized tetragonal zirconia polycrystal (Y-TZP, henceforth simply, zirconia) has been used as an articulation material in either monolithic or composite form in total hip arthroplasty (THA). Over 600,000 monolithic zirconia and 3.8 million zirconia-toughened alumina matrix (ZTA) composite femoral heads have so far been implanted worldwide. In the history of their clinical use, it was found that environmental exposure of ceramic implants including a metastable zirconia phase could trigger tetragonal-to-monoclinic (t-m) phase transformation in the human body, which is a corrosive effect induced by polar water molecules on the Zr-O-Zr bonds [1]. This phenomenon, known as low-temperature degradation (LTD) aging of zirconia, might potentially lead to in-vivo cracking/fracture of the ceramic component [2,3] and enhanced wear in polyethylene counterparts [4,5] as a consequence of the crystal volume expansion, shear strain, and surface roughening. Since the resistance to LTD can greatly affect the implant durability and long-term success of THA [1,6], a practical and clinically relevant simulation of its time-dependent behavior is essential.

Accelerated hydrothermal aging has long been one of the most important and widely accepted quality control tests for simulating LTD in zirconia-containing ceramics for medical applications [1,7–10]. The aging protocol is standardized by the International Organization for Standardization (ISO) 13356 or 6474-2, and American Society for Testing and Materials (ASTM) F2345 as a steam sterilization procedure at 134 °C under a pressure of 0.2 MPa in autoclave. According to a thermal activation energy of medical grade Y-TZP, an extrapolation of the hydrothermal effect under body temperature (37 °C) provides a theoretical estimation that 1 h in-vitro at 134 °C under 0.2 MPa has the same effect as 2 ∼ 4 yrs in-vivo [1,7–10]. ASTM F2345 specifies that 1 h aging ≈ 2 yrs in-vivo. The estimation of 1 h aging ≈ 4 yrs in-vivo is often adopted in ISO-based studies, which has been considered so far as a reasonable prediction of LTD degradation in-vivo. Chevalier [7] showed an excellent agreement between prediction at 37 °C from the accelerated tests and experimental measurements in two zirconia head retrievals after short implantation periods (4.5 and 8 yrs in unlabeled and Prozyr^® heads [St. Gobain Desmarquest, France]). Nevertheless, potential limitations for interpreting such encouraging results may lie in no information of the examined head surface regions (e.g., polar/equatorial region, main-wear/transition-wear/non-wear region), and also in a lack of long-term follow-up including more severe and complex effects of in-vivo tribology and chemistry on LTD.

The main purpose of the present study is twofold: (1) to evaluate the long-term consistency between the prediction from standardized tests and the actual measurements from various surface regions of long-term zirconia head retrievals; and (2) to clarify the long-term effects of in-vivo mechanics/tribology on the LTD progression upon service in the body environment.

2. Materials and methods

2.1. Surgically removed zirconia heads

We investigated five long-term retrievals (13.1–18.4 yrs) of 28-mm-sized zirconia (Prozyr^®) femoral heads (Fig. 1) coupled with ultra-high molecular weight polyethylene acetabular liners (ArCom^®, Biomet Inc., Warsaw, IN, USA). All retrievals were free of surgical malposition. The examined heads did not contain alumina dopant and were subjected to hot isostatic pressing (HIPing). Note also that none of them belonged to the tunnel furnace lots resulting in the worldwide recall in 2001 [2,11]. The mean grain size and yttria stabilizer content in the Prozyr^® material were 0.5 μm and 3 mol%, respectively.

Fig. 1.

Photographs of surgically removed zirconia femoral heads after long-term in-vivo service. The black arrows indicate metal transfer.

Table 1

Information about patients and zirconia femoral heads analyzed in this study

Head number	Head size (cm)	Age at surgery	Gender	Implant side	BMI (kg/m²)	Time in-vivo (yrs)	Reason for removal
#1	28	56	Female	Right	22.3	13.1	PE wear
#2	28	57	Female	Right	20.4	14.8	PE wear
#3	28	57	Female	Left	21.5	16.6	PE wear
#4	28	51	Female	Right	21.5	17.7	PE wear
#5	28	55	Female	Right	20.9	18.4	PE wear

BMI: body mass index; PE: polyethylene.

Primary and revision THA were performed at the same single institution. The stated indication for revision surgery was polyethylene liner wear. The available information on patients and retrievals is summarized in Table1. In order to investigate the effect of implantation period on LTD, we matched gender and reason of revision (polyethylene wear), and also homogenized body mass index (BMI) and age at surgery.

All retrievals were inspected for evidence of surface wear, scratches, or retention of machining marks using an optical microscope (magnification ×100). Surface wear was characterized by loss of the original machining marks from the manufacturing process. The following four zones were identified on head surfaces through their microscopic observation (Fig. 2a–b): (I) non-wear zone (NWZ); (II) medial wear zone (MWZ); (III) polar wear zone (PWZ); and (IV) superior wear zone (SWZ). Head alignment was confirmed and marked at the time of revision surgery. PWZ and SWZ were located within the so-called main-wear zone.

Fig. 2.

(a) Schematic of the analyzed locations on zirconia head retrievals. NWZ, MWZ, PWZ, and SWZ represent non-wear, medial-wear, polar-wear, and superior-wear zones; (b) Optical micrographs (magnification 100×) taken on the surfaces of zirconia head retrievals after different times of in-vivo service. The original linear machining marks in NWZ are highlighted by the white dotted lines. The black arrows, yellow arrowheads, and white circles indicate scratches, pits, and metallic stains, respectively.

2.2. Confocal Raman microspectroscopy

In each selected zone of retrievals, surface monoclinic volume percentage (transformed percentage) and residual stress in tetragonal phase were analyzed by confocal Raman microspectroscopy. 488-nm Ar-ion laser (GLG3103, Showa Optronics Co., Ltd., Tokyo, Japan) was irradiated onto head surfaces to excite Raman spectra, whose power was set at 30 mW. The laser beam was focused through an Olympus MPlanFLN microscope objective of 100× magnification and N.A. = 0.80 (Tokyo, Japan). The focal spot size was ∼1 μm, and a diameter of pinhole aperture was set at 100 μm to filter out-of-focus emission light through confocal configuration. Raman spectra were detected at room temperature (25 ± 2 °C) using the spectrometer (MS3504i SOL instruments Ltd., Minsk, Republic of Belarus) equipped with a thermoelectrically-cooled (−90 °C) charge-coupled device (CCD) camera (1024 × 256 pixels; iDus DU-420A-BR-DD, Andor Technology, Belfast, Northern Ireland, UK). A holographic grating (1800 lines/mm) was used to resolve the spectra. The spectrometer, the CCD detector, and a three-coordinate (xyz) micro-stage were controlled via Raman Scope ver.2.01 software (Lambda Vision Inc., Kanagawa, Japan). Individual Raman spectra were collected in 5 sec, and then averaged over three successive measurements for each location measured. A Raman spectral mapping was performed at ∼9-μm steps within the area of 100 × 100 μm². A total of 432 spectra were collected in each zone.

The recorded data were exported from Raman Scope into Labspec 3 software (HORIBA Jobin Yvon SAS, Lille, France) for the spectral deconvolution. A mixed Gaussian/Lorentzian curve fit was applied to deconvolute into sub-bands, and peak intensities and positions were analyzed. The monoclinic volume percentage (V _m) was determined by peak intensities at 150, 180, and 190 cm⁻¹ using the Katagiri equation [12]. The residual stress in tetragonal phase (𝜎_t) was calculated from peak shift of 260 cm⁻¹ based on the piezo-spectroscopic (PS) theory [13], and the reference peak position in each selected surface was taken from the corresponding locations of the unused (as-received) Prozyr^® head. These computations were made by the following equations: $\begin{eqnarray}\displaystyle \text{} & \text{} & \displaystyle V_{m}=100\times \frac{0.5(I_{m}^{180}+I_{m}^{190})}{2.2I_{t}^{150}+0.5(I_{m}^{180}+I_{m}^{190})}\end{eqnarray}$ (1) $\begin{eqnarray}\displaystyle \text{} & \text{} & \displaystyle {\sigma}_{t}=\frac{{\Delta}V_{t}^{260}}{{\Pi}}\end{eqnarray}$ (2) where I _t and I _m represent the peak intensities of the Raman bands from tetragonal and monoclinic phases. 𝛥𝜈_t and 𝛱 represent the spectral shift of Raman peak from tetragonal phase and its PS coefficient [13]. The wavenumbers are identified by the respective superscripts.

The V _m and 𝜎_t values on each retrieval surface were compared to those obtained by hydrothermal aging simulation, which we previously conducted according to ASTM and ISO criteria (henceforth referred to as SIM-I [1 h aging = 2 yrs in-vivo] and SIM-II [1 h aging = 4 yrs in-vivo]) [1,7–10]. In addition, the environmentally and tribologically driven phase transformation ( ${\Delta}V_{m}^{E}$ , ${\Delta}V_{m}^{T}$ ) and residual stresses ( ${\Delta}{\sigma}_{t}^{E}$ , ${\Delta}{\sigma}_{t}^{T}$ ) were approximated by means of the following Eqs ((3))–((6)). These equations were based on the hypothesis that wear zones were subjected to both in-vivo environment and mechanical-tribological damages while the non-wear zones were affected by the environment alone: $\begin{eqnarray}\displaystyle \text{} & \text{} & \displaystyle {\Delta}V_{m}^{E}=V_{m}^{N}-V_{m}^{0}\end{eqnarray}$ (3) $\begin{eqnarray}\displaystyle \text{} & \text{} & \displaystyle {\Delta}V_{m}^{T}=V_{m}^{W}-V_{m}^{N}\end{eqnarray}$ (4) $\begin{eqnarray}\displaystyle \text{} & \text{} & \displaystyle {\Delta}{\sigma}_{t}^{E}={\sigma}_{t}^{N}-{\sigma}_{t}^{0}\end{eqnarray}$ (5) $\begin{eqnarray}\displaystyle \text{} & \text{} & \displaystyle {\Delta}{\sigma}_{t}^{T}={\sigma}_{t}^{W}-{\sigma}_{t}^{N}\end{eqnarray}$ (6) where the superscripts, 0, N, and W, represent the as-received, non-wear, and wear zone surfaces, respectively.

2.3. Statistical analysis

A two-tailed Student’s t test (with Welch’s correction) was performed with the aid of Graphpad Prism software, version 6.05 (GraphPad software, Inc., San Diego, CA, USA), to test for statistically significant differences of transformed percentages and residual stresses among the selected surface area. The statistical differences in these comparisons were considered significant at the p < 0.05 level.

3. Results and discussion

3.1. Transformed percentage and residual stress in head retrievals

Figure 2b shows the optical micrographs (x100) collected in each zone of the surface of the long-term zirconia head retrievals after different implantation periods ranging from 13.1 to 18.4 yrs (Head #1–#5). The original linear machining marks from the manufacturing process were observed in NWZ (suggesting surface intactness in the area), whereas they were conspicuously eliminated from the wear zones. The machining marks were still present in MWZ, but they were much less clear than in NWZ. SWZ and PWZ belonging to the main-wear zone showed more disrupted surfaces with multiple curvilinear scratches, pits, and metal contamination. Surface metal stains were observed in Head #1, #2, #4, and #5 (Figs 1 and 2b). Although scratches and small pits existed also in MWZ, their extents and concentrations were much milder than SWZ and PWZ. Surface uplifts related to volume expansion after t-m transformation were distributed over the three wear zones (SWZ, PWZ, MWZ), and isolated clusters of monoclinic grains were sparsely observed to a much lesser extent in NWZ.

Fig. 3.

Comparisons of average Raman spectra of zirconia femoral heads obtained from the unused surface and the selected surface of Head #5. Lowercase characters t and m in the spectra indicate the peak positions assigned to tetragonal and monoclinic phase zirconia, respectively.

Figure 3 shows the representative Raman spectra recorded on the surfaces of the as-received head and retrieval #5 (18.4 yrs). Raman bands on the unused one indicate predominant spectral contributions from tetragonal phase in zirconia (150, 260, 320, 465, 605, 640 cm⁻¹). Bands assigned to the monoclinic polymorph (180, 190, 220, 300, 330, 345, 380, 475, 535, 560, 615, 635 cm⁻¹) clearly appeared in the retrieval, and their peak intensities were more pronounced in the wear zones than in the NWZ. The most notable increases of their intensities were found in the PWZ, indicating the most transformed region.

The peak intensity values were quantitatively converted into monoclinic volume percentages ( $V_{m}^{0}$ , V _m). In the unused state, the surface regions corresponding to the NWZ, MWZ, PWZ, and SWZ in the retrievals originally had $V_{m}^{0}=2.8$ , 2.9, 3.0, and 3.0%, respectively, and no statistically significant differences were noted among these regions. As compared to the unused surfaces, significant t − m transformation was detected on all investigated zones of retrievals ( $V_{m}>V_{m}^{0}$ , p < 0.0001). Since articulating surfaces were subjected to in-vivo environment plus mechanical-tribological damages, the V _m values in MWZ, PWZ, and SWZ were significantly higher than those in NWZ (p < 0.0001, Fig. 4a). According to the spectroscopic findings, we demonstrated that the in-vivo service produced tensile residual stress in tetragonal phase (𝜎_t) on retrieval surfaces (Fig. 4b). The tensile magnitudes increased as phase transformation increased. Note that PWZ always exhibited the highest degrees of t − m transformation and tensile stress, suggesting that the effect of mechanical-tribological damages was greatest among the examined regions.

Fig. 4.

Average monoclinic volume percentage (a) and residual tensile stress in tetragonal phase (b) as measured on the selected surface zones in zirconia head retrievals. ^∗∗∗∗ represents p < 0.0001.

3.2. Consistency between predicted and actual LTD behaviors

In this section we discuss the consistency in each zone between the prediction from the standardized tests and the actual measurements from retrievals after long-term service. Retrieval data are plotted for comparison together with simulated aging data (SIM-I: 1 h = 2yrs; SIM-II: 1 h = 4 yrs), which were previously obtained according to the same Raman algorithms [10] (Fig. 5a–b). The V _m profile in NWZ is in good agreement with SIM-II, whereas the V _m plots in SWZ and MWZ are closer to SIM-I. However, PWZ shows much higher V _m than both simulations. On the other hand, the degree of data consistency in 𝜎_t is relatively lower than that in V _m. All wear zones exhibit a much higher level of residual tension than predicted. In addition, despite the good agreement of V _m between NWZ and SIM-II, higher tensile magnitude was detected in NWZ whose values were within the range of SIM-I and II. The above discrepancies may be partly associated with hoop stress from surgical impaction to firmly fix stem trunnion into head bore. However, since NWZ has a stepwise stress increase with in-vivo years, other factors could be responsible for surface residual tension. Another conceivable reason could be the exposure to different environments (i.e., synovial fluid vs. water alone). It was previously pointed out that environmentally driven off-stoichiometry and strain accumulation could generate as a consequence of biochemical interactions between oxide ceramic surfaces and water molecules/synovial fluid including some substitutable cations (e.g., Ca²⁺, Mg²⁺) [14–16]. The formation of some point defects such as oxygen vacancies, and substitutional (aliovalent) cations can be expected in ceramic head retrievals, resulting in lattice strain fields (of both elastic and plastic nature) in the very neighborhood of the surfaces [14–16]. In this context, SIM-II may somewhat underestimate the environmental stress in NWZ, whereas SIM-I clearly overestimate as LTD prediction in NWZ. Nevertheless, both simulations are not accurately predictive for the long-term wear surfaces since there were conjunctions of environmental and mechanical-tribological effects. The latter is discussed in the next section.

Fig. 5.

Comparisons of the LTD kinetics are made between real-time in-vivo aging and in-vitro accelerated simulations at 134 °C (SIM-I: 1 h = 2 yrs; SIM-II: 1 h = 4 yrs [from Arita et al. [10]]) by co-plotting their transformed percentages (a) and the tensile stresses developed in the tetragonal phase (b).

3.3. Biotribological contribution to LTD

The worn surfaces of our zirconia retrievals aged far faster than predicted by ASTM and ISO-recommended aging, suggesting that biotribology had a strong effect on LTD. Given that such contribution is defined as subtraction of analyzed data in NWZ from those in each wear zone, it can be quantified as shown in Tables2 and3. As compared to environmental transformation ( ${\Delta}V_{m}^{E}$ ), tribological transformation ( ${\Delta}V_{m}^{T}$ ) was 1.3–4.3 times higher in MWZ, 2.4–6.5 times higher in PWZ, and 1.9–5.3 times higher in SWZ. On the other hand, tribological stress ( ${\Delta}{\sigma}_{t}^{T}$ ) was 0.8–2.3 times higher in MWZ, 1.4–3.3 times higher in PWZ, and 1.1–2.6 times higher in SWZ than environmental stress ( ${\Delta}{\sigma}_{t}^{E}$ ).

Table 2
Environmental vs. tribological phase transformation in each selected zone of zirconia femoral head retrievals after long-term in-vivo service

Time in-vivo (yrs) Environmental transformation Tribological transformation

MWZ PWZ SWZ

${\Delta}V_{m}^{E}$ (%) ${\Delta}V_{m}^{T}$ (%) ${\Delta}V_{m}^{T}/{\Delta}V_{m}^{E}$ ${\Delta}V_{m}^{T}$ (%) ${\Delta}V_{m}^{T}/{\Delta}V_{m}^{E}$ ${\Delta}V_{m}^{T}$ (%) ${\Delta}V_{m}^{T}/{\Delta}V_{m}^{E}$

13.3 6.5 28.2 4.3 42.4 6.5 34.6 5.3

14.8 8.5 31.8 3.7 45.0 5.3 35.8 4.2

16.6 7.9 24.8 3.1 43.7 5.5 32.0 4.0

17.7 14.6 19.4 1.3 35.5 2.4 27.8 1.9

18.4 14.4 33.4 2.3 48.6 3.4 44.5 3.1

Time in-vivo (yrs)	Environmental transformation	Tribological transformation
13.3	6.5	28.2	4.3	42.4	6.5	34.6	5.3
14.8	8.5	31.8	3.7	45.0	5.3	35.8	4.2
16.6	7.9	24.8	3.1	43.7	5.5	32.0	4.0
17.7	14.6	19.4	1.3	35.5	2.4	27.8	1.9
18.4	14.4	33.4	2.3	48.6	3.4	44.5	3.1

Table 3

Environmental vs. tribological stress in each selected zone of zirconia femoral head retrievals after long-term in-vivo service

Time in-vivo (yrs)	Environmental stress	Tribological stress
		MWZ		PWZ		SWZ
	${\Delta}{\sigma}_{t}^{E}$ (MPa)	${\Delta}{\sigma}_{t}^{T}$ (MPa)	${\Delta}{\sigma}_{t}^{T}/{\Delta}{\sigma}_{t}^{E}$	${\Delta}{\sigma}_{t}^{T}$ (MPa)	${\Delta}{\sigma}_{t}^{T}/{\Delta}{\sigma}_{t}^{E}$	${\Delta}{\sigma}_{t}^{T}$ (MPa)	${\Delta}{\sigma}_{t}^{T}/{\Delta}{\sigma}_{t}^{E}$
13.3	72.6	168.6	2.3	241.5	3.3	187.5	2.6
14.8	82.5	162.8	2.0	256.6	3.1	180.4	2.2
16.6	102.4	163.7	1.6	216.9	2.1	162.5	1.6
17.7	133.3	106.8	0.8	193.0	1.4	150.0	1.1
18.4	144.3	149.3	1.0	253.7	1.8	209.1	1.4

Possible biotribological contributions to aging acceleration could be envisaged as follows: (1) loading, wear, third-body entrapment, scratching, and the related frictional heating; (2) mechanical impingement and shocks due to separation of bearing surfaces; (3) tribochemical reactions with hydroxyls and aliovalent cations from synovial fluid; and (4) frictionally-driven metal transfer. The pressure-temperature effects in wear zones under polyethylene cups play a role in phase transformation. In zirconia-polyethylene bearings, surface temperature can reach up to 99 °C during head-cup friction due to the low thermal conductivity of zirconia [17]. However, according to a study by Brown et al. [18], zirconia-polyethylene wear reproduced in hip simulator did not accelerate phase transformation, whereas the magnitudes of transformation in wear surfaces of Prozyr^® head retrievals were greater at 3–8 yrs follow-up than predicted by an in-vitro aging study. Such discrepancy has been considered as a lack of biological processes in hip simulator including homeostasis, metabolism, and catabolism [19]. As a result, precipitation of thermally degraded proteins in hip simulator protected bearing surfaces from wear [18,19], leading to less transformation compared to the in-vivo use. Interarticular impingement and shocks due to separation can be a trigger for phase transformation of zirconia in ceramic-on-ceramic bearings [20]. Since our heads were coupled with polyethylene liners, which act as shock absorbers, joint separation is thought to contribute to deformation of polyethylene liner rather than to the LTD process in zirconia [21]. We do not believe that shocks, especially those in the polar region induced by surgical impaction, notably affected polymorphic transformation in zirconia. Mechanochemical interactions resulting in off-stoichiometry and lattice plasticity may occur in the worn ceramic surfaces, as unveiled by high-resolution cathodoluminescence spectroscopy [14–16]. Moreover, recent studies found that zirconia destabilization and grain boundary embrittlement were induced by transition metal contamination on the surfaces of ZTA hip implants [22–25]. In our series, metal transfer was found in SWZ and PWZ of Head #1, #2, #4, #5, and in MWZ of Head #1, #2 (Fig. 1 and 2b), and it thus might have affected the LTD kinetics.

In the above contexts, there still seems to be some critical factors missing in the current in-vitro aging protocols. Nevertheless, latest experimental attempts were made to combine environmental and mechanical factors. Gremillard et al. [26] have suggested an improved protocol for in-vitro aging test of metastable ceramics, which includes the effects of loading and wear in evaluating hydrothermal stability of hard-on-soft bearings. Perrichon et al. [20] have further suggested including mechanical shocks due to joint separation for simulating hard-on-hard bearings. These new protocols potentially represent an improvement of the present standard, although it might not be exhaustive in including the chemical factors (i.e., (3)--(4)).

4. Conclusion

Our study evaluated the clinically relevant magnitudes of long-term biotribological contribution to LTD in zirconia as compared to environmental exposure alone. The present findings could contribute to future developments of more clinically-relevant and accurate in-vitro simulations. The main conclusions of the present study can be summarized as follows:

Levels of monoclinic fraction and tensile stress were far higher in wear zones (especially in the head polar region) than in non-wear zones, suggesting that biotribology had a strong effect on aging kinetics of zirconia.

Long-term aging behavior in non-wear zones was in agreement with the experimental prediction estimated as 1 h aging at 134 °C = 4 years in-vivo.

The current in-vitro aging protocols were not accurately predictive for the long-term wear surfaces since there were conjunctions of environmental and tribological transformation triggers.

Tribologically induced transformation and tensile stress were up to 6.5-times and 3.3-times higher than the environmentally induced ones.

Wear/scratching, frictional heating, tribochemical reactions, and metal transfer could have constituted additional triggers in accelerating polymorphic transformation at the zirconia head surfaces.

Footnotes

Acknowledgements

This work was financially supported in part by a Grant-in-Aid for Scientific Research (C) Project (grant no. 18K04756) from the Japanese Society for the Promotion of Science (JSPS).

Conflict of interest

The authors declare no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

References

Clarke

I.C.

, Metastable nature of zirconia femoral heads from a 20-year perspective of clinical and simulator wear studies, Semin. Arthroplast. 17 (2006), 165–178.

Masonis

J.L.

Bourne

R.B.

Ries

M.D.

McCalden

R.W.

Salehi

and Kelman

D.C.

, Zirconia femoral head fractures: A clinical and retrieval analysis, J. Arthroplast. 19 (2004), 898–905.

Hummer III

C.D.

Rothman

R.H.

and Hozack

W.J.

, Catastrophic failure of modular zirconia-ceramic femoral head components after total hip arthroplasty, J. Arthroplast. 10 (1995), 848–850.

Liang

Kawanabe

Ise

Iida

and Nakamura

, Polyethylene wear against alumina and zirconia heads in cemented total hip arthroplasty, J. Arthroplast. 22 (2007), 251–257.

Fernandez-Fairen

Blanco

Murcia

Sevilla

and Gil

F.J.

, Aging of retrieved zirconia femoral heads, Clin. Orthop. Relat. Res. 462 (2007), 122–129.

Yoshitomi

Shikata

Ito

Nakayama

and Nakamura

, Manufacturers affect clinical results of THA with zirconia heads: A systematic review, Clin. Orthop. Relat. Res. 467 (2009), 2349–2355.

Chevalier

, What future for zirconia as a biomaterial? Biomater. 27 (2006), 535–543.

Chevalier

Gremillard

and Deville

, Low-temperature degradation of zirconia and implications for biomedical implants, Annu. Rev. Mater. Res. 37 (2007), 1–32.

Pezzotti

Saito

Takahashi

Fukatsu

and Sugano

, Surface topology of advanced alumina/zirconia composite femora head as compared with commercial femoral heads made of monolithic zirconia, J. Am. Ceram. Soc. 94 (2011), 945–950.

10.

Arita

Takahashi

Pezzotti

Shishido

Masaoka

Sano

and Yamamoto

, Environmental stability and residual stresses in zirconia femoral head for total hip arthroplasty: In vitro aging versus retrieval studies, Biomed Res Int. 2015 (2015), 638502.

11.

Clarke

I.C.

Manaka

Green

D.D.

Williams

Pezzotti

Kim

Y.H.

Ries

Sugano

Sedel

Delauney

Nissan

B.B.

Donaldson

and Gustafson

G.A.

, J. Bone. Jt. Surg. Am. 85 (2003), 73–84.

12.

Katagiri

Ishida

Ishitani

and Masaki

, Direct determination by a Raman microprobe of transformation zone size in Y₂O₃-containing tetragonal ZrO₂ polycrystals, in: Science and Technology of Zirconia III, S. Somiya, N. Yamamoto and H. Yanagida (eds), Advances in Ceramics, Vol. 24, Am. Ceram. Soc., Columbus, 1988, pp. 537--544.

13.

Pezzotti

and Porporati

A.A.

, Raman spectroscopic analysis of phase-transformation and stress patterns in zirconia hip joints, J. Biomed. Opt. 9 (2004), 372–384.

14.

Pezzotti

Munisso

M.C.

Porporati

A.A.

and Lessnau

, On the role of oxygen vacancies and lattice strain in the tetragonal to monoclinic transformation in alumina/zirconia composites and improved environmental stability, Biomater 31 (2010), 6901–6908.

15.

Takahashi

Zhu

Sugano

and Pezzotti

, On the role of oxygen vacancies, aliovalent ions and lattice strain in the in vivo wear behavior of alumina hip joints, J. Mech. Behav. Biomed. Mater. 4 (2011), 993–1003.

16.

Fukatsu

Leto

Zhu

Sugano

and Pezzotti

, Kinetics and the role of off-stoichiometry in the environmentally driven phase transformation of commercially available zirconia femoral heads, Acta Biomater. 8 (2012), 1639–1647.

17.

and McKellop

, Frictional heating of bearing materials tested in a hip joint wear simulator, Proc. Inst. Mech. Eng. H. 211 (1997), 101–108.

18.

Brown

S.S.

Green

D.D.

Pezzotti

Donaldson

T.K.

and Clarke

I.C.

, Possible triggers for phase transformation in zirconia hip balls, J. Biomed. Mater. Res. B. Appl. Biomater. 85 (2008), 444–452.

19.

Essner

and Wang

, Tribological assessment of UHMWPE in the hip, in: UHMWPE Biomaterials Handbook, Second Edition: Ultra-High Molecular Weight Polyethylene in Total Joint Replacement and Medical Devices Kurtz

S.M.

(ed.), Academic Press, Burlington, (2009), pp. 369–379.

20.

Perrichon

Reynard

Gremillard

Chevalier

Farizon

and Geringer

, A testing protocol combining shocks, hydrothermal ageing and friction, applied to Zirconia Toughened Alumina (ZTA) hip implants, J. Mech. Behav. Biomed. Mater. 65 (2017), 600–608.

21.

Williams

Butterfield

Stewart

Ingham

Stone

and Fisher

, Wear and deformation of ceramic-on-polyethylene total hip replacements with joint laxity and swing phase microseparation, Proc. Inst. Mech. Eng. H. 217 (2003), 147–153.

22.

Boffelli

Doimo

Marin

Puppulin

Zhu

Sugano

Clarke

I.C.

and Pezzotti

, Chemically driven tetragonal-to-monoclinic transformation in retrieved ZTA femoral heads from dual mobility hip implants, J. Mech. Behav. Biomed. Mater. 56 (2016), 195–204.

23.

Bal

B.S.

Zhu

Zanocco

Marin

Sugano

McEntire

B.J.

and Pezzotti

, Reconciling in vivo and in vitro kinetics of the polymorphic transformation in zirconia-toughened alumina for hip joints: I. Phenomenology, Mater. Sci. Eng. C Mater. Biol. Appl. 72 (2017), 252–258.

24.

Pezzotti

Zhu

Zanocco

Marin

Sugano

McEntire

B.J.

and Bal

B.S.

, Reconciling in vivo and in vitro kinetics of the polymorphic transformation in zirconia-toughened alumina for hip joints: II. Theory, Mater. Sci. Eng. C Mater. Biol. Appl. 71 (2017), 446–451.

25.

Pezzotti

Bal

B.S.

Zanocco

Marin

Sugano

McEntire

B.J.

and Zhu

, Reconciling in vivo and in vitro kinetics of the polymorphic transformation in zirconia-toughened alumina for hip joints: III. Molecular scale mechanisms, Mater. Sci. Eng. C Mater. Biol. Appl. 71 (2017), 552–557.

26.

Gremillard

Martin

Zych

Crosnier

Chevalier

Charbouillot

Sainsot

Espinouse

and Aurelle

J.L.

, Combining ageing and wear to assess the durability of zirconia-based ceramic heads for total hip arthroplasty, Acta Biomater. 9 (2013), 7545–7555.