- Email: [email protected]
Improved Resistance to Neck-Liner Impingement in Second-Generation Highly Crosslinked Polyethylene—The Role of Vitamin E and Crosslinks
The Journal of Arthroplasty xxx (2016) 1e7
Contents lists available at ScienceDirect
The Journal of Arthroplasty journal homepage: www.arthroplastyjournal.org
Original Article
Improved Resistance to Neck-Liner Impingement in SecondGeneration Highly Crosslinked PolyethylenedThe Role of Vitamin E and Crosslinks Yasuhito Takahashi, PhD a, b, *, Toshiyuki Tateiwa, MD, PhD a, Giuseppe Pezzotti, PhD c, Takaaki Shishido, MD, PhD a, Toshinori Masaoka, MD, PhD a, Kengo Yamamoto, MD, PhD a a b c
Department of Orthopedic Surgery, Tokyo Medical University, Tokyo, Japan Department of Bone and Joint Biomaterial Research, Tokyo Medical University, Tokyo, Japan Ceramic Physics Laboratory, Kyoto Institute of Technology, Kyoto, Japan
a r t i c l e i n f o
a b s t r a c t
Article history: Received 6 February 2016 Received in revised form 26 April 2016 Accepted 23 May 2016 Available online xxx
Background: Radiation crosslinking of ultrahigh molecular weight polyethylene (UHMWPE) results in the reduced tensile strength and fracture toughness as an expense of dramatic increase in the wear resistance. Clinical rim fracture has been reported due to neck-liner impingement on a first-generation highly crosslinked UHMWPE acetabular component. The objective of this study was to investigate whether a second-generation, vitamin Eeblended highly crosslinked UHMWPE possesses the improved impingement resistance. Methods: Cyclic impingement testing was performed in a variety of UHMWPE acetabular components (vitamin E free or blended, noncrosslinked or highly crosslinked, and GUR1050 or GUR1020) with the same design specification. The kinematics used to reproduce the neck-liner impingement was a uniaxial fatigue compression in concert with an axial rotational torque. After the test, the geometry and morphological changes were characterized by coordinate measuring machine, scanning electron microscopy, and confocal Raman microspectroscopy. Results: A total of 300-kGy irradiated and annealed GUR1050 liner resulted in a significant geometry change and microcracks on the rim surface after the test. However, regardless of the similar level of crosslinking, much less damage was noted in the 300-kGy irradiated GUR1050 liner blended with vitamin E at a concentration of 3000 ppm. On the other hand, vitamin Eeblended noncrosslinked GUR1050 exhibited an extensive microscopic fibrillation and folding on the impinged surface. Conclusion: These results suggest that vitamin Eeblending into UHMWPE has compensated the negative effect of toughness decrease induced by radiation crosslinking. We concluded that the coexistence of vitamin E and crosslinks can restrain impingement damage more effectively than either of them. © 2016 Elsevier Inc. All rights reserved.
Keywords: UHMWPE acetabular component neck-liner impingement vitamin E blend crosslinks
Impingement between a neck of femoral stem and a rim of acetabular liner may result in a significant risk of implant failures in total hip arthroplasty (THA). The repetitive rim impingement against ultrahigh molecular weight polyethylene (UHMWPE)
One or more of the authors of this paper have disclosed potential or pertinent conflicts of interest, which may include receipt of payment, either direct or indirect, institutional support, or association with an entity in the biomedical field which may be perceived to have potential conflict of interest with this work. For full disclosure statements refer to http://dx.doi.org/10.1016/j.arth.2016.05.049. * Reprint requests: Yasuhito Takahashi, PhD, Department of Orthopedic Surgery, Tokyo Medical University, 6-7-1 Nishishinjuku, Shinjuku-ku, Tokyo 160-0023, Japan. http://dx.doi.org/10.1016/j.arth.2016.05.049 0883-5403/© 2016 Elsevier Inc. All rights reserved.
acetabular liners is known to contribute to the occurrence of plastic deformation, wear debris release (rim, articular, and backside wear), delamination, fatigue crack propagation, peripheral fracture, and femoral-head dislocation [1-9]. The evidence of neck-liner impingement is commonly identified by the macroscopic changes in the rim geometry of UHMWPE liners, and these in vivo prevalences were reported as 27%-84% in THA [1-6]. The first-generation radiation crosslinked and melted UHMWPE liners can generally entail a greater risk of cracking and fracture at the rim sites after impingement than the conventional (noncrosslinked or low crosslinked) ones [10,11]. The formation of intermolecular covalent crosslinking bonds in UHMWPE results in the reduced tensile strength and fracture toughness as an expense
2
Y. Takahashi et al. / The Journal of Arthroplasty xxx (2016) 1e7
of dramatic increase in the wear resistance. The ductile-to-brittle transition, thus, proceeds in UHMWPE with increasing doses of radiation [10-12]. In addition, post-irradiation melting, which increases oxidative stability by quenching the trapped free radicals, contributes to a further reduction of the mechanical properties due to a loss of the crystallinity [11]. In principle, the mechanical properties including the impingement resistance is associated with the microstructural features such as molecular weight, crosslink density, phase balance between crystalline and noncrystalline regions, degree of anisotropy, and presence of antioxidants (such as vitamin E) as well as their homogeneity within UHMWPE. The strategies for equilibrating among wear resistance, oxidative stability, and mechanical properties varied among orthopedic device makers. The starting resins of UHMWPE can be found to be a significant factor affecting the property balance between wear resistance and impact strength. The calcium stearateefree UHMWPE resin, GUR 1050 or GUR 1020 (Celanese, Inc, Florence, KY), which has a molecular weight of 5.5-6.0 or 3.5 million g/mol, are used in a majority of the modern acetabular liners. GUR 1050 was found to have a greater resistance to wear but less impact strength and tensile mechanical properties than GUR 1020 [13]. However, the multiple effects of the above structural features on physical behaviors are quite complex. In particular, intense recent attention has been given to the mechanical performance of the second-generation highly crosslinked UHMWPE stabilized with vitamin E, which is nowadays recognized as one of the new standards in THA [14,15]. In the present study, to update our understanding of neck-liner impingement, we set the following 3 questions: (1) Does crosslinking increase the peripheral rim damage of the acetabular liner? (2) Does vitamin E-blending mitigate the impingement lesion? (3) How are the combined (competitive) effect of vitamin E blending and crosslinking against impingement? For this purpose, cyclic rim impingement testing was performed for a variety of UHMWPE acetabular components (vitamin E free or blended, noncrosslinked or highly crosslinked, and GUR 1050 or 1020) with the same design specification. After the test, the geometry and morphological changes on their rims were nondestructively characterized by coordinate measuring machine, scanning electron microscopy (SEM), and confocal Raman microspectroscopy (CRS). Materials and Methods Preparation of UHMWPE Acetabular Liners All UHMWPE specimens were manufactured either from GUR 1050 or 1020. The resin powders were consolidated via direct compression molding into the semi-final shape of acetabular liner. Five different groups of the test liners, GUR1050, GUR1050-E, GUR1050-XL, GUR1050-EXL, and GUR1020-XL (n ¼ 4 samples for each group) were prepared in the same design and dimension (28-mm inner diameter and 7.5-mm thickness). The abbreviations, -E, -XL, and -EXL stand for the presence of blended vitamin E, crosslinks, and both of them in the microstructure, respectively. GUR1050-EXL is commercially known as BLEND-E XL produced by Teijin Nakashima Medical Co, Ltd (Okayama, Japan), which was approved by the Japanese Pharmaceuticals and Medical Devices Agency. BLEND-E XL was clinically introduced in 2013 for THA. In each type of the materials, 3 of the 4 samples were used for the impingement test, and the remaining one sample served as a control material for the geometrical and morphological characterizations, that is, its unused state before the testing was analyzed in terms of the rim geometry, surface percentages of crystalline, amorphous, and intermediate phase volumes. The methodological details of the analyses are given in the forthcoming sections.
Vitamin E (dl-aetocopherol, Eisai Co Ltd, Japan) was blended with GUR1050 resin at a concentration of 3000 ppm (0.3 wt/wt %) in liquid form before direct compression molding. On the other hand, crosslinking was achieved in direct molded forms via electron beam irradiation (10 MeV) in vacuum with a total dose of 300 kGy for GUR1050-XL and GUR1050-EXL and of 90 kGy for GUR1020XL. Since the e-beam can deposit quite a bit of energy into the UHMWPE structure, which likely leads to some temperature rise, the material surfaces was kept at <50 C with adopting sequential irradiation steps of 30 kGy. Being this temperature far below the melting point of UHMWPE (z150 C), no thermal effects are expected on the microstructure. After the irradiation, the materials were annealed in vacuum at 110 C for 72 hours and then machined/ polished to obtain the final shape of acetabular liners. It should be noted that the presence of vitamin E inhibits crosslinking reactions among UHMWPE molecules by scavenging some of the primary free radicals generated by irradiation and forming vitamin E radicals [16-18]. However, it was previously demonstrated that the vitamin Eeinduced loss of crosslinking can be minimized to some extent by optimizing radiation dose and subsequent thermal treatment [17,18]. The crosslink density of GUR1050-EXL was confirmed to be an almost equivalent level as GUR1050-XL according to the result of their gel contents (95.6% vs 96.6%) [18]. For the previously mentioned reasons, a 300-kGy irradiation was conducted to GUR1050-EXL which is a much higher dose than the general levels for highly crosslinked UHMWPE liners, that is, 50-100 kGy. GUR1050-XL was irradiated at the same dose as a comparative benchmark although it nearly reaches a plateau of crosslink density above 100 kGy [19]. Cyclic Impingement Testing Cyclic impingement was performed to examine the mechanical strength of the liner specimens on an electrodynamic axialtorsional materials testing machine (ElectroPuls E10000, Instron Corp, Norwood, MA). In general, assuming an appropriately placed implant, anterior and posterior impingement tend to occur with a prosthetic hip mainly in flexion and internal rotation and in extension and external rotation, respectively [2]. In such circumstances, the femoral neck does not impinge simply in a direction perpendicular to the acetabular rim surface, and accordingly the multiple-stress fields including compression and shear stresses can generate at the neck-rim contact. Thus, we reproduced the kinematics used for the neck-liner impingement testing as a uniaxial fatigue compression in concert with an axial rotational torque (Fig. 1A-B). During this testing, the acetabular rims can be exposed to impact shock, fatigue compression, and shear forces. The room temperature was controlled at 23.6 ± 2.5 C throughout the test. The used femoral heads, stems, and acetabular shells were made of cobalt chrome alloy. The 28-mm femoral heads were locked onto the stems and placed into the acetabular liners housed in the shells. The femoral stems were set at the angle of 23 against the acetabular rims. The acetabular shells were fixed in a custom-made jig by using 3 screws (Fig. 1C). After each testing, the stem necks were exchanged to exclude the possible influence of the roughness changes induced by the surface damages. The cyclic compressive load of 19-190 N (stress ratio R ¼ 0.1) was applied on the acetabular rims through the necks of the femoral stems at a frequency of 1 Hz (Fig. 1D). The peak load was determined as a maximum compression value which did not induce the neck-shell impingement in all the specimen groups. On the other hand, the rotational torque was generated by swinging the stem necks on the rims at a frequency of 1 Hz, and its rotational angle was set at the range of 0 -10 (Fig. 1D). These movements
Y. Takahashi et al. / The Journal of Arthroplasty xxx (2016) 1e7
3
Fig. 1. Photographs illustrating the loading apparatus (A), the sample setup (B), and the fixation jig (C) for cyclic neck-liner impingement. The input waveforms of uniaxial compression force (P) and the neck swing angle (q) were shown in the (D) (19 P 190 [N], and 0q 10 [deg]). The impingement scar with the size of z5.5 8.5 mm2 was formed at the acetabular rim, as surrounded by the red border in the (E). The yellow circle indicates the location where the maximum damage can be expected to occur.
allowed the necks to drag across the liner rims with the peak load of the cyclic fatigue. The testing was carried out for a total of 100,000 cycles in each specimen. In this experimental configuration, maximum damage can be generated nearby the center of the impingement scars within the chamfer surfaces of the acetabular rims (Figs. 1E and 2A-B).
Surface Analyses by Coordinate Measuring Machine and SEM Surface geometry assessment of the acetabular rims was performed at room temperature (20 ± 2 C) by using a coordinate measuring machine (PRISMO navigator; Zeiss, Oberkochen, Germany). As shown in Figure 2A, the peripheral geometry of the liners were probed using a ruby stylus with a tip radius of 1.5 mm (scan rate ¼ 3 mm/s; accuracy ¼ ±0.9 mm). A total of 6804 points were measured with a sampling of 0.2-mm step for each liner. The maximum depth of impingement scars located at each chamfer surface (maximum geometric displacement, Dmax) was analyzed by comparing and subtracting the geometry profiles recorded before and after the impingement test (Fig. 2B). This analysis was conducted by using a commercial software (Calypso; Zeiss). In addition, the chamfer surfaces were visually examined by SEM (JSM-6010LV, JEOL Ltd, Tokyo, Japan). Before the SEM observation, the specimens were fixed on aluminum disks using a carbon tapes and subsequently sputter-coated by platinum using an auto fine coater (JEC-3000FC, JEOL Ltd). SEM images were collected at an accelerated voltage of 5 kV.
Determination of Surface Percentages of Crystalline, Amorphous, and Intermediate Phase Volumes by CRS The microstructure of UHMWPE consists of the following 3 phases: (1) an ordered crystalline phase, (2) a disordered liquid-like amorphous phase, and (3) an anisotropic intermediate phase with a molecular arrangement lying between the above 2 regions. The volume percentages of the each phase were nondestructively and quantitatively determined on the chamfer surfaces of the acetabular rims by CRS. This measurement was performed before and after the impingement. The surface percentages of crystalline,
Fig. 2. Three-dimensional geometric characterization using coordinate measuring machine (CMM) and its stylus probe with a ruby ball tip (A). The computational analysis of maximum depth of impingement scars located at each chamfer surface, Dmax (B). The peripheral geometry profile examined in impinged liner specimen was compared and subtracted from that in unused specimen.
4
Y. Takahashi et al. / The Journal of Arthroplasty xxx (2016) 1e7
amorphous, and intermediate phase volumes were referred to as ac, aa, and ai, respectively. The CRS analyses for all the specimens were performed at room temperature (25 ± 2 C) by means of a Raman microprobe spectrometer (MS3504i, SOL Instruments Ltd, Minsk, Republic of Belarus) equipped with a thermoelectrically cooled (90 C) charge-coupled device camera (1024 256 pixels; iDus DU-420ABR-DD, Andor Technology, Belfast, Northern Ireland, United Kingdom). The excitation source was a 488-nm Ar-ion laser (GLG3103, Showa Optronics Co, Ltd, Tokyo, Japan) yielding a power of 30 mW. In the impinged liners, Raman spectra were measured nearby the centers of the impingement scars within the chamfer surfaces. The focal spot size on the specimen surfaces was ~1 mm via a microscope objective (magnification 100, numerical aperture 0.80, working distance 3.4 mm, LMPlanFL-N, Olympus Co, Ltd, Tokyo, Japan). A diameter of pinhole aperture was set as 100 mm to filter out-of-focus emission light (ie, Raman scattered light from the liner subsurfaces) through the confocal configuration. All the spectra were acquired in backscattering geometry with a spectral resolution of ~2.0 cm1 achieved by an 1800 grooves/mm grating. The spectrometer, the charge-coupled device detector, and a 3coordinate (xyz) microstage were controlled via Raman Scope version 1.64 software (Lambda Vision Inc, Kanagawa, Japan). Individual spectra were collected in 5 seconds and then averaged over 3 successive measurements for each point. A Raman spectral mapping was performed at 5-mm steps within the area of 75 75 mm2. A total of 1024 different points (¼256 points/map 4 maps) were analyzed around the center of the impinged areas. The recorded data were exported from Raman Scope into Labspec 3 software (HORIBA Jobin Yvon SAS, Lille, France) for the spectral fitting. A mixed Gaussian/Lorentzian curve fit was applied for the deconvolution of the recorded spectra into subbands, and subsequently, the integrated intensities were calculated from the deconvoluted subbands. The each-phase percentage was averaged among the 1024 points for each liner specimen, and computations for them were made according to the following equations [20-23]:
ac ¼
I1418 0:46ðI1296 þ I1310 Þ
(1)
aa ¼
I1080 0:79ðI1296 þ I1310 Þ
(2)
ai ¼ 1 ðac þ aa Þ
(3)
Results All the tested UHMWPE acetabular liners showed clear geometric displacements on their rims which represent the local occurrence of plastic deformation, wear, and/or cracking induced by the repetitive impingement. Figure 3 shows the results of the mean Dmax values analyzed in each liner group. GUR1050-XL liner had the largest rim displacement, and its mean Dmax value was 15% greater than GUR1050 liner (P ¼ .02, d ¼ 3.06). As explained in the previous section, since GUR1050-XL and 1050-EXL had an almost equivalent level of crosslink density [18], their data comparison can emphasize the mechanical role of vitamin E within the crosslinked structure. This comparison indicated a 28% reduction in the mean Dmax value due to the presence of 3000-ppm vitamin E (P ¼ .0477, d ¼ 2.31). Nevertheless, no significant difference was noted in the average Dmax between the noncrosslinked groups of GUR1050 and 1050-E (P ¼ .5072, d ¼ 0.6). The smallest mean displacement was found in GUR1020-XL liner group, but there was no statistically significant difference between GUR1020-XL and 1050-EXL (P ¼ .3911, d ¼ 0.79). Figure 4A-E show the SEM micrographs at the centers of impingement scars within the chamfer surfaces. Extensive microscopic fibrillation and surface folding were observed in the noncrosslinked groups, and the original machining marks totally disappeared at their impinged chamfers (Fig. 4A-B). In GUR1050XL, machining marks partially remain within the chamfer, and less fibrillation was observed compared to the noncrosslinked liners. However, its surface was extensively cracked and folded (Fig. 4C). On the other hand, no fibrillation and folding were observed in GUR1050-EXL and 1020-XL whose machining marks were still visible in the both surfaces. Figure 5A shows the percent crystallinity, ac, on the chamfer surfaces of each liner group before and after the impingement. In the unused state, the ac of GUR1050-XL and 1050-EXL was about 10% greater than the other 3 liners. In the comparisons before and after the test, the statistically significant increases occurred in the mean ac of GUR1050, 1050-E, and 1050-XL groups (P ¼ .0004, .0004, and .0244; d ¼ 5.89, 6.74, and 2.39). There were no significant ac increases in GUR 1050-EXL and 1020-XL (P ¼ .0917 and .1908; d ¼ 1.43 and 1.04). Figure 5B-C show the percent noncrystalline phases (aa and ai) on the each chamfer surface. No statistically significant differences were observed in the mean aa between the unused and impinged surfaces of GUR1050, 1050-E, 1050-XL, and 1020-XL (P ¼ .8134, .2326, .0972, and .6531; d ¼ 0.18, 1.03, 1.50, and 0.36). However, the significant amorphization proceeded only in
where I is the integral intensity of the Raman band whose wave number is identified by the subscript.
Statistical Analysis A 2-tailed Student t test was performed with the aid of Graphpad Prism software, version 6.05 (GraphPad software, Inc, San Diego, CA), to test for statistically significant differences of the analyzed data (Dmax, ac, aa, and ai) between unused and impinged specimens of each material group and also among each impinged specimen. The statistical differences in the previously mentioned comparisons were considered significant at the P < .05 level. The effect size of each comparison was determined by Cohen's d, which represents a small effect if between 0.2 and 0.3, a medium effect if around 0.5, and a large effect if between 0.8 and infinity [24].
Fig. 3. Maximum depth of impingement scars on the chamfer surfaces (Dmax) analyzed by CMM in the each impinged UHMWPE specimen. The symbol, *, represents P < .05. UHMWPE, ultrahigh molecular weight polyethylene.
Y. Takahashi et al. / The Journal of Arthroplasty xxx (2016) 1e7
5
Fig. 4. Representative scanning electron microscopy micrographs at the centers of impingement scars on the chamfer surfaces in the each impinged UHMWPE specimen (original magnification 650). The white arrows indicate the surface folding and fibrillation, and the black arrow indicates cracking (A-C). No surface folding, fibrillation, and cracks were observed in GUR1050-EXL and GUR1020-XL liner groups (D and E).
GUR1050-EXL (P ¼ .0018; d ¼ 4.71). On the other hand, there were significant ai reductions due to impingement in GUR1050, 1050-E, 1050-XL, and 1050-EXL (P ¼ .0003, .0009, .0267, and .0042; d ¼ 6.10, 5.67, 2.27, and 3.50). Note that, any statistically significant structural alterations were not introduced in any phases of GUR1020-XL. The mean differences in each phase percentage between the unused and impinged surfaces (Dac, Daa, and Dai) were directly compared in Figure 6A-C. The highest percentage of the Dac was found in GUR1050, whereas the lowest was in GUR1050-EXL. The significantly less increase in ac was found in the crosslinked groups than the noncrosslinked. The Dac in GUR1050-XL was 57% lower than that in GUR1050 (P ¼ .0307, d ¼ 2.67). GUR1050-EXL showed the 72% lower value of Dac compared to GUR1050-E (P ¼ .0144; d ¼ 3.38). On the other hand, although 3000-ppm vitamin E blend did not statistically significantly affect the crystallinity changes according to the comparison of GUR1050 vs 1050-E and of GUR1050-XL vs 1050-EXL (P ¼ .5024 and .3769; d ¼ 0.6 and 0.81), the mean Dac values of GUR1050-E and -EXL were 11% and 42% less than GUR1050 and 1050-XL, respectively. In addition, the mean Daa value was enhanced by coexistence of vitamin E and crosslinks, while the extent of the ai reduction was not significantly affected by vitamin E blend as well as crosslinks.
Discussion Under the current impingement simulation, the UHMWPE acetabular rims can simultaneously undergo multiple damage modes such as elastic/plastic deformation, wear, and microcracking in response to impact shock, fatigue compression, and repetitive shear stress through the femoral necks. Therefore, higher tensile strength, fatigue, toughness, and wear properties of the specimens can be translated to greater resistance against the impingement. The investigated liners exhibited clear geometric changes due to the impingement. The small standard deviations of each measurement in Figures 3, 5, and 6 indicate good repeatability, and all the data comparisons with a P < .05 have an effect size of d > 0.8 (large effect size). Thus, the sample size of the present study can be considered to be adequate in satisfying the test validity at the selected level of significance (P < .05). The ac increases in the UHMWPE surfaces can be mainly attributed to the following 2 reasons: (1) solid-state recrystallization induced by a local plastic flow and (2) wear-induced denudation of the subsurface regions with higher crystallinity. The former mechanism involves the mechanically assisted local densification and defect reduction in noncrystalline phases as a result of decrease
Fig. 5. Comparisons of the 3-phase volume percentages on the chamfer surfaces in the each UHMWPE specimen before and after the cyclic impingement. Percent crystallinity, ac (A), percent amorphous phase, aa (B), and percent intermediate phase, ai (C). The symbol, *, represents P < .05.
6
Y. Takahashi et al. / The Journal of Arthroplasty xxx (2016) 1e7
Fig. 6. Differences in ac, aa, and ai between the unused and the impinged chamfer surfaces in the each UHMWPE liner specimen (A-C). The symbol, *, represents P < .05.
in interlamellar distance [25-28]. On the other hand, the latter mechanism is based on the experimental fact that there is a significant loss of crystallinity in the near-surface region (within ~100-mm depth from the surface) due to surface machining and polishing performed at the end of the manufacturing process [23,28]. As shown in Figure 4D, the surface evidence of cracking was not observed in GUR1050-EXL, and the presence of the remaining machining marks at the rim surface indicates an absence of severe wear. Thus, plastic deformation primarily contributed to the overall displacement of its rim geometry. Note that, only GUR1050-EXL liner group did not experience the intermediate-tocrystalline transition but the intermediate-to-amorphous transition (Fig. 5A-C). This fact implies that the mechanically assisted recrystallization process was hindered by the presence of vitamin E in GUR1050-EXL, and such crystallization blocking phenomenon was also observed in the same liner under an uniaxial static compression (on 10% plastic deformation of the original thickness) [23]. There have been several reports regarding the altered mechanical behaviors due to the presence of vitamin E. Oral et al [29] showed that vitamin Eediffused highly crosslinked UHMWPE (GUR 1050 resin) had significantly higher fatigue resistance, ultimate tensile strength, yield strength, and elongation at break than the conventional (g-sterilized virgin). Wolf et al [30] showed a reduced elastic modulus and ultimate tensile strength, and increased impact strength and breaking elongation in the 8000-ppm vitamin Eeblended UHMWPE (GUR 1020 resin), and Turner et al [31] also reported the similar mechanical behavior in the 30,000-ppm vitamin E-blended UHMWPE (GUR 1050 resin). In more recent studies, Takahashi et al [23, 32] demonstrated that vitamin E blending (3000 ppm) as well as diffusing/homogenization [32] provoked the increased chain mobility of highly crosslinked UHMWPE surfaces (GUR 1050 resin) in response to a severe uniaxial compression. It is obvious that 300-kGy radiation crosslinking in GUR1050-XL adversely affected the impingement resistance due to the toughness decrease, leading to the surface microcracking (Figs. 3 and 4C). Nevertheless, as far as the impingement resistance was concerned, such negative effect of toughness decrease due to crosslinking was successfully compensated by the 3000-ppm vitamin E blend. Although vitamin E was blended in GUR1050-E at the same concentration, its surface exhibited the severe wear (Fig. 4B) and the significant ac increase (Fig. 6A). These behaviors can be related to the previous mechanism (2), that is, the inner structure was denuded by wear due to the lack of crosslinks. As shown in Figures 3 and 6A, there were no statistically significant differences in the geometric and morphological changes between GUR1050 and GUR1050-E. In the previous contexts, their primary damage modes may not be cracking but surface wear and plastic deformation. Nevertheless, crack propagation and fracture were not clinically a rare event in noncrosslinked virgin UHMWPE acetabular liners because of the oxidation embrittlement in the rim [4]. In other words, owing to
the significantly higher oxidative stability in GUR1050-E [18,30], the potential risk of the rim fracture may be lower than the virgin GUR1050 during the long service life in vivo. It should be noted that, except for the differences in the chemical properties against the in vivo oxidation mechanisms (ie, long-term exposure of body fluid, and lipid absorption [33,34]), GUR1020-XL showed in its nonoxidized state a comparably- low level of rim displacement and surface wear to GUR1050-EXL (Figs. 3 and 4D-F). Furthermore, no statistically significant phase changes were confirmed in GUR1020-XL (Fig. 5A-C). Although crosslinking was achieved by 90-kGy e-beam irradiation in GUR1020-XL, the Dmax value was significantly lower than the noncrosslinked GUR1050 and 1050-E (P ¼ .0380 and 0.0189). These data may indicate a beneficial aspect of using GUR 1020 resin as an acetabular liner, that is, superior mechanical properties than GUR 1050 resin [13]. In the recent clinical report, the rim fracture of a vitamin Eediffused highly crosslinked liner (GUR 1050 resin) occurred in a case with an excessive combined anteversion (59 ), leading to posterior impingement [35]. This might shift the attention toward the vitamin Eestabilized highly crosslinked GUR1020 liner, and this type of the liner is actually becoming more popular among the main commercial brands [36]. Therefore, it would be important to keep our eyes on the clinical impacts of the different starting resins as well as balance of crosslink density and vitamin E concentration in UHMWPE. Our study has limitations that may temper the extent of the simulated impingement lesions compared to those in the actual in vivo environment. First, the testing environment was in air at room temperature (23.6 ± 2.5 C), not in synovial fluid at body temperature. Thus, oxidation mechanism related to lipid absorption [34] was not included. The testing and subsequent characterizations were completed in a relatively short period (within 6 months) of time after their preparation, and the specimens were vacuum packed except at the time of the experiments. The timedependent oxidative degradation during the shelf storage in air can be negligible amount. Second, we tested using only one type of compression profile including the restricted peak force (up to 190 N). In the previous study by Oral et al [29], a peak load of 125 lb (~556 N) was used for the rim impingement testing, based on a torque of 100 in lb, a moment necessary to cause dislocation of the hip joint [37]. In the present study, however, the neck-shell impingement was avoided by restricting the maximum compressive load in the investigated type of the prosthetic design to achieve adequate repeatability of the experiment. Thus, our loading condition was not severe enough to trigger hip dislocation but was within a range of force magnitudes which can possibly occur in vivo. Third, the used kinematics may not cover all the possible modes of the in vivo impingement. Finally, a single rim design, thickness, and dimension were examined. The rim geometry relative to that of a metal neck can affect the level of the impingement damage. Nevertheless, since the main purpose of this study was to compare differences in the mechanical responses among the
Y. Takahashi et al. / The Journal of Arthroplasty xxx (2016) 1e7
different materials, the reproducible methodology as well as the identical neck-liner design was essential for the test. Despite aforementioned limitations, our results showed that the vitamin E blend improved the resistance to cyclic impingement on highly crosslinked UHMWPE acetabular liner. Acknowledgments The authors thank Nakashima Medical for preparing the commercial and noncommercial samples. Special thanks to Dr Keita Uetsuki and Mr Yuta Osaka for their technical supports in the impingement testing and for many helpful discussions. References 1. Wiadrowski TP, McGee M, Cornish BL, et al. Peripheral wear of Wagner resurfacing hip arthroplasty acetabular components. J Arthroplasty 1991;6:103. 2. Yamaguchi M, Akisue T, Bauer T, et al. The spatial location of impingement in total hip arthroplasty. J Arthroplasty 2000;15(3):305. 3. Shon WY, Baldini T, Peterson MG, et al. Impingement in total hip arthroplasty: a study of retrieved acetabular components. J Arthroplasty 2005;20(4):427. 4. Birman MV, Noble PC, Conditt MA, et al. Cracking and impingement in ultrahigh-molecular-weight polyethylene acetabular liners. J Arthroplasty 2005;20(7):87. 5. Usrey MM, Noble PC, Rudner LJ, et al. Does neck/liner impingement increase wear of ultrahigh-molecular-weight polyethylene liners? J Arthroplasty 2006;21(6):65. 6. Tanino H, Harman MK, Banks SA, et al. Association between dislocation, impingement, and articular geometry in retrieved acetabular polyethylene cups. J Orthop Res 2007;29:465. 7. Salehi A, Yamamoto K, Wong P, et al. Mechanical performance of a bipolar design. Transactions of the 55th Orthopaedic Research Society, Las Vegas (NV); 2009. p. 2305. 8. Furmanski J, Anderson M, Bal S, et al. Clinical fracture of cross-linked UHMWPE acetabular liners. Biomaterials 2009;30(29):5572. 9. Marchetti E, Krantz N, Berton C, et al. Component impingement in total hip arthroplasty: frequency and risk factors. A continuous retrieval analysis series of 416 cup. Orthop Traumatol Surg Res 2011;97(2):127. 10. Pruitt LA. Deformation, yielding, fracture and fatigue behavior of conventional and highly crosslinked ultra high molecular weight polyethylene. Biomaterials 2005;26(8):905. 11. Muratoglu OK, Bragdon CR. Highly cross-linked and melted UHMWPE. In: Kurtz SM, editor. Uhmwpe Biomaterials Handbook, 3rd ed.: Ultra-High Molecular Weight Polyethylene in Total Joint Replacement and Medical Devices. Oxford: William Andrew; 2015. p. 264. 12. Suarez JCM, de Biasi RS. Effect of gamma irradiation on the ductile-to-brittle transition in ultra-high molecular weight polyethylene. Polym Degrad Stabil 2003;82(2):221. 13. Cybo J, Maszybrocka J, Barylski A, et al. Resistance of UHMWPE to plastic deformation and wear and the possibility of its enhancement through modification by radiation. J Appl Polym Sci 2012;125(6):4188. 14. Bracco P, Oral E. Vitamin E-stabilized UHMWPE for total joint implants. Clin Orthop Relat Res 2011;469(8):2286. 15. ASTM International. ASTM F 2695e12. Standard specification for ultra-high molecular weight polyethylene powder blended with alpha-tocopherol (vitamin E) and fabricated forms for surgical implant applications. West Conshohocken (PA): ASTM International; 2012.
7
16. Oral E, Godleski Beckos C, Malhi AS, et al. The effects of high dose irradiation on the cross-linking of vitamin E-blended ultrahigh molecular weight polyethylene. Biomaterials 2008;29:3557. 17. Kawasaki T, Hamada D, Tajima K, et al. Annealing stabilizes vitamin E radicals and accelerates crosslinking reaction in electron-beam-irradiated dl-alpha-tocopherol-blended ultra high molecular weight polyethylene. J Bone Joint Surg Br 2012;94(SUPP XL):90. 18. Uetsuki K, Sugimoto T, Turner AC, et al. Simultaneous improvement in the mechanical strength and oxidative resistance of vitamin E blended UHMWPE. Transactions of the 57th Orthopaedic Research Society, San Francisco (CA); 2012. p. 1073. 19. Muratoglu H, Bragdon CR, O'Connor DO, et al. Unified wear model for highly crosslinked ultra-high molecular weight polyethylenes (UHMWPE). Biomaterials 1999;20:1463. 20. Strobl GR, Hagedorn W. Raman spectroscopic method for determining the crystallinity of polyethylene. J Polym Sci Polym Phys Ed 1978;16:1181. 21. Glotin M, Mandelkern L. A Raman spectroscopic study of the morphological structure of the polyethylenes. Colloid Polym Sci 1982;260:182. 22. Mutter R, Stille W, Strobl G. Transition regions and surface melting in partially crystalline polyethylene: a Raman spectroscopic study. J Polym Sci B Polym Phy 1993;31:99. 23. Takahashi Y, Yamamoto K, Pezzotti G. Effects of vitamin-E blending on plastic deformation mechanisms of highly cross-linked ultra-high molecular weight polyethylene (HXL-UHMWPE) in total hip arthroplasty. Acta Biomater 2015;15:227. 24. Cohen J. A power primer. Psychol Bull 1992;112(1):155. 25. Bowden PB, Young RJ. Deformation mechanisms in crystalline polymers. J Mater Sci 1974;9(12):2034. 26. Kim JM, Locker R, Rutledge GC. Plastic deformation of semicrystalline polyethylene under extension, compression, and shear using molecular dynamics simulation. Macromolecules 2014;47(7):2515. 27. Lin L, Argon AS. Structure and plastic deformation of polyethylene. J Mater Sci 1994;29(2):294. 28. Takahashi Y, Shishido T, Yamamoto K, et al. Mechanisms of plastic deformation in highly cross-linked UHMWPE for total hip components e the molecular physics viewpoint. J Mech Behav Biomed Mater 2015;42:43. 29. Oral E, Christensen SD, Malhi AS, et al. Wear resistance and mechanical properties of highly cross-linked, ultrahigh-molecular weight polyethylene doped with vitamin E. J Arthroplasty 2006;21(4):580. 30. Wolf C, Krivec T, Blassnig J, et al. Examination of the suitability of alphatocopherol as a stabilizer for ultra-high molecular weight polyethylene used for articulating surfaces in joint endoprostheses. J Mater Sci Mater Med 2002;13(2):185. 31. Turner A, Okubo Y, Teramura S, et al. The antioxidant and non-antioxidant contributions of vitamin E in vitamin E blended ultra-high molecular weight polyethylene for total knee replacement. J Mech Behav Biomed Mater 2014;31:21. 32. Takahashi Y, Yamamoto K, Shishido T, et al. Strain-induced microstructural rearrangement in ultra-high molecular weight polyethylene for hip joints: a comparison between conventional and vitamin E-infused highly-crosslinked liners. J Mech Behav Biomed Mater 2014;31:31. 33. Muratoglu OK, Wannomae KK, Rowell SL, et al. Ex vivo stability loss of irradiated and melted UHMWPE. J Bone Joint Surg Am 2010;92(17):2809. 34. Oral E, Ghali BW, Neils AL, et al. A new mechanism of oxidation in UHMWPE caused by squalene absorption. J Biomed Mater Res B Appl Biomater 2012;100(3):742. 35. Beecher B, Huddleston J, Rowell S, et al. Material properties of a highly anteverted vitamin E-stabilized polyethylene liner after sixteen months in vivo. JBJS Case Connector 2014;4(3):e57. 36. Gigante A, Bottegoni C, Ragone V, et al. Effectiveness of vitamin-E-doped polyethylene in joint replacement: a literature review. J Funct Biomater 2015;6:889. 37. Scifert CF, Brown TD, Pedersen DR, et al. A finite element analysis of factors influencing total hip dislocation. Clin Orthop Relat Res 1998;355:152.
Contents lists available at ScienceDirect
The Journal of Arthroplasty journal homepage: www.arthroplastyjournal.org
Original Article
Improved Resistance to Neck-Liner Impingement in SecondGeneration Highly Crosslinked PolyethylenedThe Role of Vitamin E and Crosslinks Yasuhito Takahashi, PhD a, b, *, Toshiyuki Tateiwa, MD, PhD a, Giuseppe Pezzotti, PhD c, Takaaki Shishido, MD, PhD a, Toshinori Masaoka, MD, PhD a, Kengo Yamamoto, MD, PhD a a b c
Department of Orthopedic Surgery, Tokyo Medical University, Tokyo, Japan Department of Bone and Joint Biomaterial Research, Tokyo Medical University, Tokyo, Japan Ceramic Physics Laboratory, Kyoto Institute of Technology, Kyoto, Japan
a r t i c l e i n f o
a b s t r a c t
Article history: Received 6 February 2016 Received in revised form 26 April 2016 Accepted 23 May 2016 Available online xxx
Background: Radiation crosslinking of ultrahigh molecular weight polyethylene (UHMWPE) results in the reduced tensile strength and fracture toughness as an expense of dramatic increase in the wear resistance. Clinical rim fracture has been reported due to neck-liner impingement on a first-generation highly crosslinked UHMWPE acetabular component. The objective of this study was to investigate whether a second-generation, vitamin Eeblended highly crosslinked UHMWPE possesses the improved impingement resistance. Methods: Cyclic impingement testing was performed in a variety of UHMWPE acetabular components (vitamin E free or blended, noncrosslinked or highly crosslinked, and GUR1050 or GUR1020) with the same design specification. The kinematics used to reproduce the neck-liner impingement was a uniaxial fatigue compression in concert with an axial rotational torque. After the test, the geometry and morphological changes were characterized by coordinate measuring machine, scanning electron microscopy, and confocal Raman microspectroscopy. Results: A total of 300-kGy irradiated and annealed GUR1050 liner resulted in a significant geometry change and microcracks on the rim surface after the test. However, regardless of the similar level of crosslinking, much less damage was noted in the 300-kGy irradiated GUR1050 liner blended with vitamin E at a concentration of 3000 ppm. On the other hand, vitamin Eeblended noncrosslinked GUR1050 exhibited an extensive microscopic fibrillation and folding on the impinged surface. Conclusion: These results suggest that vitamin Eeblending into UHMWPE has compensated the negative effect of toughness decrease induced by radiation crosslinking. We concluded that the coexistence of vitamin E and crosslinks can restrain impingement damage more effectively than either of them. © 2016 Elsevier Inc. All rights reserved.
Keywords: UHMWPE acetabular component neck-liner impingement vitamin E blend crosslinks
Impingement between a neck of femoral stem and a rim of acetabular liner may result in a significant risk of implant failures in total hip arthroplasty (THA). The repetitive rim impingement against ultrahigh molecular weight polyethylene (UHMWPE)
One or more of the authors of this paper have disclosed potential or pertinent conflicts of interest, which may include receipt of payment, either direct or indirect, institutional support, or association with an entity in the biomedical field which may be perceived to have potential conflict of interest with this work. For full disclosure statements refer to http://dx.doi.org/10.1016/j.arth.2016.05.049. * Reprint requests: Yasuhito Takahashi, PhD, Department of Orthopedic Surgery, Tokyo Medical University, 6-7-1 Nishishinjuku, Shinjuku-ku, Tokyo 160-0023, Japan. http://dx.doi.org/10.1016/j.arth.2016.05.049 0883-5403/© 2016 Elsevier Inc. All rights reserved.
acetabular liners is known to contribute to the occurrence of plastic deformation, wear debris release (rim, articular, and backside wear), delamination, fatigue crack propagation, peripheral fracture, and femoral-head dislocation [1-9]. The evidence of neck-liner impingement is commonly identified by the macroscopic changes in the rim geometry of UHMWPE liners, and these in vivo prevalences were reported as 27%-84% in THA [1-6]. The first-generation radiation crosslinked and melted UHMWPE liners can generally entail a greater risk of cracking and fracture at the rim sites after impingement than the conventional (noncrosslinked or low crosslinked) ones [10,11]. The formation of intermolecular covalent crosslinking bonds in UHMWPE results in the reduced tensile strength and fracture toughness as an expense
2
Y. Takahashi et al. / The Journal of Arthroplasty xxx (2016) 1e7
of dramatic increase in the wear resistance. The ductile-to-brittle transition, thus, proceeds in UHMWPE with increasing doses of radiation [10-12]. In addition, post-irradiation melting, which increases oxidative stability by quenching the trapped free radicals, contributes to a further reduction of the mechanical properties due to a loss of the crystallinity [11]. In principle, the mechanical properties including the impingement resistance is associated with the microstructural features such as molecular weight, crosslink density, phase balance between crystalline and noncrystalline regions, degree of anisotropy, and presence of antioxidants (such as vitamin E) as well as their homogeneity within UHMWPE. The strategies for equilibrating among wear resistance, oxidative stability, and mechanical properties varied among orthopedic device makers. The starting resins of UHMWPE can be found to be a significant factor affecting the property balance between wear resistance and impact strength. The calcium stearateefree UHMWPE resin, GUR 1050 or GUR 1020 (Celanese, Inc, Florence, KY), which has a molecular weight of 5.5-6.0 or 3.5 million g/mol, are used in a majority of the modern acetabular liners. GUR 1050 was found to have a greater resistance to wear but less impact strength and tensile mechanical properties than GUR 1020 [13]. However, the multiple effects of the above structural features on physical behaviors are quite complex. In particular, intense recent attention has been given to the mechanical performance of the second-generation highly crosslinked UHMWPE stabilized with vitamin E, which is nowadays recognized as one of the new standards in THA [14,15]. In the present study, to update our understanding of neck-liner impingement, we set the following 3 questions: (1) Does crosslinking increase the peripheral rim damage of the acetabular liner? (2) Does vitamin E-blending mitigate the impingement lesion? (3) How are the combined (competitive) effect of vitamin E blending and crosslinking against impingement? For this purpose, cyclic rim impingement testing was performed for a variety of UHMWPE acetabular components (vitamin E free or blended, noncrosslinked or highly crosslinked, and GUR 1050 or 1020) with the same design specification. After the test, the geometry and morphological changes on their rims were nondestructively characterized by coordinate measuring machine, scanning electron microscopy (SEM), and confocal Raman microspectroscopy (CRS). Materials and Methods Preparation of UHMWPE Acetabular Liners All UHMWPE specimens were manufactured either from GUR 1050 or 1020. The resin powders were consolidated via direct compression molding into the semi-final shape of acetabular liner. Five different groups of the test liners, GUR1050, GUR1050-E, GUR1050-XL, GUR1050-EXL, and GUR1020-XL (n ¼ 4 samples for each group) were prepared in the same design and dimension (28-mm inner diameter and 7.5-mm thickness). The abbreviations, -E, -XL, and -EXL stand for the presence of blended vitamin E, crosslinks, and both of them in the microstructure, respectively. GUR1050-EXL is commercially known as BLEND-E XL produced by Teijin Nakashima Medical Co, Ltd (Okayama, Japan), which was approved by the Japanese Pharmaceuticals and Medical Devices Agency. BLEND-E XL was clinically introduced in 2013 for THA. In each type of the materials, 3 of the 4 samples were used for the impingement test, and the remaining one sample served as a control material for the geometrical and morphological characterizations, that is, its unused state before the testing was analyzed in terms of the rim geometry, surface percentages of crystalline, amorphous, and intermediate phase volumes. The methodological details of the analyses are given in the forthcoming sections.
Vitamin E (dl-aetocopherol, Eisai Co Ltd, Japan) was blended with GUR1050 resin at a concentration of 3000 ppm (0.3 wt/wt %) in liquid form before direct compression molding. On the other hand, crosslinking was achieved in direct molded forms via electron beam irradiation (10 MeV) in vacuum with a total dose of 300 kGy for GUR1050-XL and GUR1050-EXL and of 90 kGy for GUR1020XL. Since the e-beam can deposit quite a bit of energy into the UHMWPE structure, which likely leads to some temperature rise, the material surfaces was kept at <50 C with adopting sequential irradiation steps of 30 kGy. Being this temperature far below the melting point of UHMWPE (z150 C), no thermal effects are expected on the microstructure. After the irradiation, the materials were annealed in vacuum at 110 C for 72 hours and then machined/ polished to obtain the final shape of acetabular liners. It should be noted that the presence of vitamin E inhibits crosslinking reactions among UHMWPE molecules by scavenging some of the primary free radicals generated by irradiation and forming vitamin E radicals [16-18]. However, it was previously demonstrated that the vitamin Eeinduced loss of crosslinking can be minimized to some extent by optimizing radiation dose and subsequent thermal treatment [17,18]. The crosslink density of GUR1050-EXL was confirmed to be an almost equivalent level as GUR1050-XL according to the result of their gel contents (95.6% vs 96.6%) [18]. For the previously mentioned reasons, a 300-kGy irradiation was conducted to GUR1050-EXL which is a much higher dose than the general levels for highly crosslinked UHMWPE liners, that is, 50-100 kGy. GUR1050-XL was irradiated at the same dose as a comparative benchmark although it nearly reaches a plateau of crosslink density above 100 kGy [19]. Cyclic Impingement Testing Cyclic impingement was performed to examine the mechanical strength of the liner specimens on an electrodynamic axialtorsional materials testing machine (ElectroPuls E10000, Instron Corp, Norwood, MA). In general, assuming an appropriately placed implant, anterior and posterior impingement tend to occur with a prosthetic hip mainly in flexion and internal rotation and in extension and external rotation, respectively [2]. In such circumstances, the femoral neck does not impinge simply in a direction perpendicular to the acetabular rim surface, and accordingly the multiple-stress fields including compression and shear stresses can generate at the neck-rim contact. Thus, we reproduced the kinematics used for the neck-liner impingement testing as a uniaxial fatigue compression in concert with an axial rotational torque (Fig. 1A-B). During this testing, the acetabular rims can be exposed to impact shock, fatigue compression, and shear forces. The room temperature was controlled at 23.6 ± 2.5 C throughout the test. The used femoral heads, stems, and acetabular shells were made of cobalt chrome alloy. The 28-mm femoral heads were locked onto the stems and placed into the acetabular liners housed in the shells. The femoral stems were set at the angle of 23 against the acetabular rims. The acetabular shells were fixed in a custom-made jig by using 3 screws (Fig. 1C). After each testing, the stem necks were exchanged to exclude the possible influence of the roughness changes induced by the surface damages. The cyclic compressive load of 19-190 N (stress ratio R ¼ 0.1) was applied on the acetabular rims through the necks of the femoral stems at a frequency of 1 Hz (Fig. 1D). The peak load was determined as a maximum compression value which did not induce the neck-shell impingement in all the specimen groups. On the other hand, the rotational torque was generated by swinging the stem necks on the rims at a frequency of 1 Hz, and its rotational angle was set at the range of 0 -10 (Fig. 1D). These movements
Y. Takahashi et al. / The Journal of Arthroplasty xxx (2016) 1e7
3
Fig. 1. Photographs illustrating the loading apparatus (A), the sample setup (B), and the fixation jig (C) for cyclic neck-liner impingement. The input waveforms of uniaxial compression force (P) and the neck swing angle (q) were shown in the (D) (19 P 190 [N], and 0q 10 [deg]). The impingement scar with the size of z5.5 8.5 mm2 was formed at the acetabular rim, as surrounded by the red border in the (E). The yellow circle indicates the location where the maximum damage can be expected to occur.
allowed the necks to drag across the liner rims with the peak load of the cyclic fatigue. The testing was carried out for a total of 100,000 cycles in each specimen. In this experimental configuration, maximum damage can be generated nearby the center of the impingement scars within the chamfer surfaces of the acetabular rims (Figs. 1E and 2A-B).
Surface Analyses by Coordinate Measuring Machine and SEM Surface geometry assessment of the acetabular rims was performed at room temperature (20 ± 2 C) by using a coordinate measuring machine (PRISMO navigator; Zeiss, Oberkochen, Germany). As shown in Figure 2A, the peripheral geometry of the liners were probed using a ruby stylus with a tip radius of 1.5 mm (scan rate ¼ 3 mm/s; accuracy ¼ ±0.9 mm). A total of 6804 points were measured with a sampling of 0.2-mm step for each liner. The maximum depth of impingement scars located at each chamfer surface (maximum geometric displacement, Dmax) was analyzed by comparing and subtracting the geometry profiles recorded before and after the impingement test (Fig. 2B). This analysis was conducted by using a commercial software (Calypso; Zeiss). In addition, the chamfer surfaces were visually examined by SEM (JSM-6010LV, JEOL Ltd, Tokyo, Japan). Before the SEM observation, the specimens were fixed on aluminum disks using a carbon tapes and subsequently sputter-coated by platinum using an auto fine coater (JEC-3000FC, JEOL Ltd). SEM images were collected at an accelerated voltage of 5 kV.
Determination of Surface Percentages of Crystalline, Amorphous, and Intermediate Phase Volumes by CRS The microstructure of UHMWPE consists of the following 3 phases: (1) an ordered crystalline phase, (2) a disordered liquid-like amorphous phase, and (3) an anisotropic intermediate phase with a molecular arrangement lying between the above 2 regions. The volume percentages of the each phase were nondestructively and quantitatively determined on the chamfer surfaces of the acetabular rims by CRS. This measurement was performed before and after the impingement. The surface percentages of crystalline,
Fig. 2. Three-dimensional geometric characterization using coordinate measuring machine (CMM) and its stylus probe with a ruby ball tip (A). The computational analysis of maximum depth of impingement scars located at each chamfer surface, Dmax (B). The peripheral geometry profile examined in impinged liner specimen was compared and subtracted from that in unused specimen.
4
Y. Takahashi et al. / The Journal of Arthroplasty xxx (2016) 1e7
amorphous, and intermediate phase volumes were referred to as ac, aa, and ai, respectively. The CRS analyses for all the specimens were performed at room temperature (25 ± 2 C) by means of a Raman microprobe spectrometer (MS3504i, SOL Instruments Ltd, Minsk, Republic of Belarus) equipped with a thermoelectrically cooled (90 C) charge-coupled device camera (1024 256 pixels; iDus DU-420ABR-DD, Andor Technology, Belfast, Northern Ireland, United Kingdom). The excitation source was a 488-nm Ar-ion laser (GLG3103, Showa Optronics Co, Ltd, Tokyo, Japan) yielding a power of 30 mW. In the impinged liners, Raman spectra were measured nearby the centers of the impingement scars within the chamfer surfaces. The focal spot size on the specimen surfaces was ~1 mm via a microscope objective (magnification 100, numerical aperture 0.80, working distance 3.4 mm, LMPlanFL-N, Olympus Co, Ltd, Tokyo, Japan). A diameter of pinhole aperture was set as 100 mm to filter out-of-focus emission light (ie, Raman scattered light from the liner subsurfaces) through the confocal configuration. All the spectra were acquired in backscattering geometry with a spectral resolution of ~2.0 cm1 achieved by an 1800 grooves/mm grating. The spectrometer, the charge-coupled device detector, and a 3coordinate (xyz) microstage were controlled via Raman Scope version 1.64 software (Lambda Vision Inc, Kanagawa, Japan). Individual spectra were collected in 5 seconds and then averaged over 3 successive measurements for each point. A Raman spectral mapping was performed at 5-mm steps within the area of 75 75 mm2. A total of 1024 different points (¼256 points/map 4 maps) were analyzed around the center of the impinged areas. The recorded data were exported from Raman Scope into Labspec 3 software (HORIBA Jobin Yvon SAS, Lille, France) for the spectral fitting. A mixed Gaussian/Lorentzian curve fit was applied for the deconvolution of the recorded spectra into subbands, and subsequently, the integrated intensities were calculated from the deconvoluted subbands. The each-phase percentage was averaged among the 1024 points for each liner specimen, and computations for them were made according to the following equations [20-23]:
ac ¼
I1418 0:46ðI1296 þ I1310 Þ
(1)
aa ¼
I1080 0:79ðI1296 þ I1310 Þ
(2)
ai ¼ 1 ðac þ aa Þ
(3)
Results All the tested UHMWPE acetabular liners showed clear geometric displacements on their rims which represent the local occurrence of plastic deformation, wear, and/or cracking induced by the repetitive impingement. Figure 3 shows the results of the mean Dmax values analyzed in each liner group. GUR1050-XL liner had the largest rim displacement, and its mean Dmax value was 15% greater than GUR1050 liner (P ¼ .02, d ¼ 3.06). As explained in the previous section, since GUR1050-XL and 1050-EXL had an almost equivalent level of crosslink density [18], their data comparison can emphasize the mechanical role of vitamin E within the crosslinked structure. This comparison indicated a 28% reduction in the mean Dmax value due to the presence of 3000-ppm vitamin E (P ¼ .0477, d ¼ 2.31). Nevertheless, no significant difference was noted in the average Dmax between the noncrosslinked groups of GUR1050 and 1050-E (P ¼ .5072, d ¼ 0.6). The smallest mean displacement was found in GUR1020-XL liner group, but there was no statistically significant difference between GUR1020-XL and 1050-EXL (P ¼ .3911, d ¼ 0.79). Figure 4A-E show the SEM micrographs at the centers of impingement scars within the chamfer surfaces. Extensive microscopic fibrillation and surface folding were observed in the noncrosslinked groups, and the original machining marks totally disappeared at their impinged chamfers (Fig. 4A-B). In GUR1050XL, machining marks partially remain within the chamfer, and less fibrillation was observed compared to the noncrosslinked liners. However, its surface was extensively cracked and folded (Fig. 4C). On the other hand, no fibrillation and folding were observed in GUR1050-EXL and 1020-XL whose machining marks were still visible in the both surfaces. Figure 5A shows the percent crystallinity, ac, on the chamfer surfaces of each liner group before and after the impingement. In the unused state, the ac of GUR1050-XL and 1050-EXL was about 10% greater than the other 3 liners. In the comparisons before and after the test, the statistically significant increases occurred in the mean ac of GUR1050, 1050-E, and 1050-XL groups (P ¼ .0004, .0004, and .0244; d ¼ 5.89, 6.74, and 2.39). There were no significant ac increases in GUR 1050-EXL and 1020-XL (P ¼ .0917 and .1908; d ¼ 1.43 and 1.04). Figure 5B-C show the percent noncrystalline phases (aa and ai) on the each chamfer surface. No statistically significant differences were observed in the mean aa between the unused and impinged surfaces of GUR1050, 1050-E, 1050-XL, and 1020-XL (P ¼ .8134, .2326, .0972, and .6531; d ¼ 0.18, 1.03, 1.50, and 0.36). However, the significant amorphization proceeded only in
where I is the integral intensity of the Raman band whose wave number is identified by the subscript.
Statistical Analysis A 2-tailed Student t test was performed with the aid of Graphpad Prism software, version 6.05 (GraphPad software, Inc, San Diego, CA), to test for statistically significant differences of the analyzed data (Dmax, ac, aa, and ai) between unused and impinged specimens of each material group and also among each impinged specimen. The statistical differences in the previously mentioned comparisons were considered significant at the P < .05 level. The effect size of each comparison was determined by Cohen's d, which represents a small effect if between 0.2 and 0.3, a medium effect if around 0.5, and a large effect if between 0.8 and infinity [24].
Fig. 3. Maximum depth of impingement scars on the chamfer surfaces (Dmax) analyzed by CMM in the each impinged UHMWPE specimen. The symbol, *, represents P < .05. UHMWPE, ultrahigh molecular weight polyethylene.
Y. Takahashi et al. / The Journal of Arthroplasty xxx (2016) 1e7
5
Fig. 4. Representative scanning electron microscopy micrographs at the centers of impingement scars on the chamfer surfaces in the each impinged UHMWPE specimen (original magnification 650). The white arrows indicate the surface folding and fibrillation, and the black arrow indicates cracking (A-C). No surface folding, fibrillation, and cracks were observed in GUR1050-EXL and GUR1020-XL liner groups (D and E).
GUR1050-EXL (P ¼ .0018; d ¼ 4.71). On the other hand, there were significant ai reductions due to impingement in GUR1050, 1050-E, 1050-XL, and 1050-EXL (P ¼ .0003, .0009, .0267, and .0042; d ¼ 6.10, 5.67, 2.27, and 3.50). Note that, any statistically significant structural alterations were not introduced in any phases of GUR1020-XL. The mean differences in each phase percentage between the unused and impinged surfaces (Dac, Daa, and Dai) were directly compared in Figure 6A-C. The highest percentage of the Dac was found in GUR1050, whereas the lowest was in GUR1050-EXL. The significantly less increase in ac was found in the crosslinked groups than the noncrosslinked. The Dac in GUR1050-XL was 57% lower than that in GUR1050 (P ¼ .0307, d ¼ 2.67). GUR1050-EXL showed the 72% lower value of Dac compared to GUR1050-E (P ¼ .0144; d ¼ 3.38). On the other hand, although 3000-ppm vitamin E blend did not statistically significantly affect the crystallinity changes according to the comparison of GUR1050 vs 1050-E and of GUR1050-XL vs 1050-EXL (P ¼ .5024 and .3769; d ¼ 0.6 and 0.81), the mean Dac values of GUR1050-E and -EXL were 11% and 42% less than GUR1050 and 1050-XL, respectively. In addition, the mean Daa value was enhanced by coexistence of vitamin E and crosslinks, while the extent of the ai reduction was not significantly affected by vitamin E blend as well as crosslinks.
Discussion Under the current impingement simulation, the UHMWPE acetabular rims can simultaneously undergo multiple damage modes such as elastic/plastic deformation, wear, and microcracking in response to impact shock, fatigue compression, and repetitive shear stress through the femoral necks. Therefore, higher tensile strength, fatigue, toughness, and wear properties of the specimens can be translated to greater resistance against the impingement. The investigated liners exhibited clear geometric changes due to the impingement. The small standard deviations of each measurement in Figures 3, 5, and 6 indicate good repeatability, and all the data comparisons with a P < .05 have an effect size of d > 0.8 (large effect size). Thus, the sample size of the present study can be considered to be adequate in satisfying the test validity at the selected level of significance (P < .05). The ac increases in the UHMWPE surfaces can be mainly attributed to the following 2 reasons: (1) solid-state recrystallization induced by a local plastic flow and (2) wear-induced denudation of the subsurface regions with higher crystallinity. The former mechanism involves the mechanically assisted local densification and defect reduction in noncrystalline phases as a result of decrease
Fig. 5. Comparisons of the 3-phase volume percentages on the chamfer surfaces in the each UHMWPE specimen before and after the cyclic impingement. Percent crystallinity, ac (A), percent amorphous phase, aa (B), and percent intermediate phase, ai (C). The symbol, *, represents P < .05.
6
Y. Takahashi et al. / The Journal of Arthroplasty xxx (2016) 1e7
Fig. 6. Differences in ac, aa, and ai between the unused and the impinged chamfer surfaces in the each UHMWPE liner specimen (A-C). The symbol, *, represents P < .05.
in interlamellar distance [25-28]. On the other hand, the latter mechanism is based on the experimental fact that there is a significant loss of crystallinity in the near-surface region (within ~100-mm depth from the surface) due to surface machining and polishing performed at the end of the manufacturing process [23,28]. As shown in Figure 4D, the surface evidence of cracking was not observed in GUR1050-EXL, and the presence of the remaining machining marks at the rim surface indicates an absence of severe wear. Thus, plastic deformation primarily contributed to the overall displacement of its rim geometry. Note that, only GUR1050-EXL liner group did not experience the intermediate-tocrystalline transition but the intermediate-to-amorphous transition (Fig. 5A-C). This fact implies that the mechanically assisted recrystallization process was hindered by the presence of vitamin E in GUR1050-EXL, and such crystallization blocking phenomenon was also observed in the same liner under an uniaxial static compression (on 10% plastic deformation of the original thickness) [23]. There have been several reports regarding the altered mechanical behaviors due to the presence of vitamin E. Oral et al [29] showed that vitamin Eediffused highly crosslinked UHMWPE (GUR 1050 resin) had significantly higher fatigue resistance, ultimate tensile strength, yield strength, and elongation at break than the conventional (g-sterilized virgin). Wolf et al [30] showed a reduced elastic modulus and ultimate tensile strength, and increased impact strength and breaking elongation in the 8000-ppm vitamin Eeblended UHMWPE (GUR 1020 resin), and Turner et al [31] also reported the similar mechanical behavior in the 30,000-ppm vitamin E-blended UHMWPE (GUR 1050 resin). In more recent studies, Takahashi et al [23, 32] demonstrated that vitamin E blending (3000 ppm) as well as diffusing/homogenization [32] provoked the increased chain mobility of highly crosslinked UHMWPE surfaces (GUR 1050 resin) in response to a severe uniaxial compression. It is obvious that 300-kGy radiation crosslinking in GUR1050-XL adversely affected the impingement resistance due to the toughness decrease, leading to the surface microcracking (Figs. 3 and 4C). Nevertheless, as far as the impingement resistance was concerned, such negative effect of toughness decrease due to crosslinking was successfully compensated by the 3000-ppm vitamin E blend. Although vitamin E was blended in GUR1050-E at the same concentration, its surface exhibited the severe wear (Fig. 4B) and the significant ac increase (Fig. 6A). These behaviors can be related to the previous mechanism (2), that is, the inner structure was denuded by wear due to the lack of crosslinks. As shown in Figures 3 and 6A, there were no statistically significant differences in the geometric and morphological changes between GUR1050 and GUR1050-E. In the previous contexts, their primary damage modes may not be cracking but surface wear and plastic deformation. Nevertheless, crack propagation and fracture were not clinically a rare event in noncrosslinked virgin UHMWPE acetabular liners because of the oxidation embrittlement in the rim [4]. In other words, owing to
the significantly higher oxidative stability in GUR1050-E [18,30], the potential risk of the rim fracture may be lower than the virgin GUR1050 during the long service life in vivo. It should be noted that, except for the differences in the chemical properties against the in vivo oxidation mechanisms (ie, long-term exposure of body fluid, and lipid absorption [33,34]), GUR1020-XL showed in its nonoxidized state a comparably- low level of rim displacement and surface wear to GUR1050-EXL (Figs. 3 and 4D-F). Furthermore, no statistically significant phase changes were confirmed in GUR1020-XL (Fig. 5A-C). Although crosslinking was achieved by 90-kGy e-beam irradiation in GUR1020-XL, the Dmax value was significantly lower than the noncrosslinked GUR1050 and 1050-E (P ¼ .0380 and 0.0189). These data may indicate a beneficial aspect of using GUR 1020 resin as an acetabular liner, that is, superior mechanical properties than GUR 1050 resin [13]. In the recent clinical report, the rim fracture of a vitamin Eediffused highly crosslinked liner (GUR 1050 resin) occurred in a case with an excessive combined anteversion (59 ), leading to posterior impingement [35]. This might shift the attention toward the vitamin Eestabilized highly crosslinked GUR1020 liner, and this type of the liner is actually becoming more popular among the main commercial brands [36]. Therefore, it would be important to keep our eyes on the clinical impacts of the different starting resins as well as balance of crosslink density and vitamin E concentration in UHMWPE. Our study has limitations that may temper the extent of the simulated impingement lesions compared to those in the actual in vivo environment. First, the testing environment was in air at room temperature (23.6 ± 2.5 C), not in synovial fluid at body temperature. Thus, oxidation mechanism related to lipid absorption [34] was not included. The testing and subsequent characterizations were completed in a relatively short period (within 6 months) of time after their preparation, and the specimens were vacuum packed except at the time of the experiments. The timedependent oxidative degradation during the shelf storage in air can be negligible amount. Second, we tested using only one type of compression profile including the restricted peak force (up to 190 N). In the previous study by Oral et al [29], a peak load of 125 lb (~556 N) was used for the rim impingement testing, based on a torque of 100 in lb, a moment necessary to cause dislocation of the hip joint [37]. In the present study, however, the neck-shell impingement was avoided by restricting the maximum compressive load in the investigated type of the prosthetic design to achieve adequate repeatability of the experiment. Thus, our loading condition was not severe enough to trigger hip dislocation but was within a range of force magnitudes which can possibly occur in vivo. Third, the used kinematics may not cover all the possible modes of the in vivo impingement. Finally, a single rim design, thickness, and dimension were examined. The rim geometry relative to that of a metal neck can affect the level of the impingement damage. Nevertheless, since the main purpose of this study was to compare differences in the mechanical responses among the
Y. Takahashi et al. / The Journal of Arthroplasty xxx (2016) 1e7
different materials, the reproducible methodology as well as the identical neck-liner design was essential for the test. Despite aforementioned limitations, our results showed that the vitamin E blend improved the resistance to cyclic impingement on highly crosslinked UHMWPE acetabular liner. Acknowledgments The authors thank Nakashima Medical for preparing the commercial and noncommercial samples. Special thanks to Dr Keita Uetsuki and Mr Yuta Osaka for their technical supports in the impingement testing and for many helpful discussions. References 1. Wiadrowski TP, McGee M, Cornish BL, et al. Peripheral wear of Wagner resurfacing hip arthroplasty acetabular components. J Arthroplasty 1991;6:103. 2. Yamaguchi M, Akisue T, Bauer T, et al. The spatial location of impingement in total hip arthroplasty. J Arthroplasty 2000;15(3):305. 3. Shon WY, Baldini T, Peterson MG, et al. Impingement in total hip arthroplasty: a study of retrieved acetabular components. J Arthroplasty 2005;20(4):427. 4. Birman MV, Noble PC, Conditt MA, et al. Cracking and impingement in ultrahigh-molecular-weight polyethylene acetabular liners. J Arthroplasty 2005;20(7):87. 5. Usrey MM, Noble PC, Rudner LJ, et al. Does neck/liner impingement increase wear of ultrahigh-molecular-weight polyethylene liners? J Arthroplasty 2006;21(6):65. 6. Tanino H, Harman MK, Banks SA, et al. Association between dislocation, impingement, and articular geometry in retrieved acetabular polyethylene cups. J Orthop Res 2007;29:465. 7. Salehi A, Yamamoto K, Wong P, et al. Mechanical performance of a bipolar design. Transactions of the 55th Orthopaedic Research Society, Las Vegas (NV); 2009. p. 2305. 8. Furmanski J, Anderson M, Bal S, et al. Clinical fracture of cross-linked UHMWPE acetabular liners. Biomaterials 2009;30(29):5572. 9. Marchetti E, Krantz N, Berton C, et al. Component impingement in total hip arthroplasty: frequency and risk factors. A continuous retrieval analysis series of 416 cup. Orthop Traumatol Surg Res 2011;97(2):127. 10. Pruitt LA. Deformation, yielding, fracture and fatigue behavior of conventional and highly crosslinked ultra high molecular weight polyethylene. Biomaterials 2005;26(8):905. 11. Muratoglu OK, Bragdon CR. Highly cross-linked and melted UHMWPE. In: Kurtz SM, editor. Uhmwpe Biomaterials Handbook, 3rd ed.: Ultra-High Molecular Weight Polyethylene in Total Joint Replacement and Medical Devices. Oxford: William Andrew; 2015. p. 264. 12. Suarez JCM, de Biasi RS. Effect of gamma irradiation on the ductile-to-brittle transition in ultra-high molecular weight polyethylene. Polym Degrad Stabil 2003;82(2):221. 13. Cybo J, Maszybrocka J, Barylski A, et al. Resistance of UHMWPE to plastic deformation and wear and the possibility of its enhancement through modification by radiation. J Appl Polym Sci 2012;125(6):4188. 14. Bracco P, Oral E. Vitamin E-stabilized UHMWPE for total joint implants. Clin Orthop Relat Res 2011;469(8):2286. 15. ASTM International. ASTM F 2695e12. Standard specification for ultra-high molecular weight polyethylene powder blended with alpha-tocopherol (vitamin E) and fabricated forms for surgical implant applications. West Conshohocken (PA): ASTM International; 2012.
7
16. Oral E, Godleski Beckos C, Malhi AS, et al. The effects of high dose irradiation on the cross-linking of vitamin E-blended ultrahigh molecular weight polyethylene. Biomaterials 2008;29:3557. 17. Kawasaki T, Hamada D, Tajima K, et al. Annealing stabilizes vitamin E radicals and accelerates crosslinking reaction in electron-beam-irradiated dl-alpha-tocopherol-blended ultra high molecular weight polyethylene. J Bone Joint Surg Br 2012;94(SUPP XL):90. 18. Uetsuki K, Sugimoto T, Turner AC, et al. Simultaneous improvement in the mechanical strength and oxidative resistance of vitamin E blended UHMWPE. Transactions of the 57th Orthopaedic Research Society, San Francisco (CA); 2012. p. 1073. 19. Muratoglu H, Bragdon CR, O'Connor DO, et al. Unified wear model for highly crosslinked ultra-high molecular weight polyethylenes (UHMWPE). Biomaterials 1999;20:1463. 20. Strobl GR, Hagedorn W. Raman spectroscopic method for determining the crystallinity of polyethylene. J Polym Sci Polym Phys Ed 1978;16:1181. 21. Glotin M, Mandelkern L. A Raman spectroscopic study of the morphological structure of the polyethylenes. Colloid Polym Sci 1982;260:182. 22. Mutter R, Stille W, Strobl G. Transition regions and surface melting in partially crystalline polyethylene: a Raman spectroscopic study. J Polym Sci B Polym Phy 1993;31:99. 23. Takahashi Y, Yamamoto K, Pezzotti G. Effects of vitamin-E blending on plastic deformation mechanisms of highly cross-linked ultra-high molecular weight polyethylene (HXL-UHMWPE) in total hip arthroplasty. Acta Biomater 2015;15:227. 24. Cohen J. A power primer. Psychol Bull 1992;112(1):155. 25. Bowden PB, Young RJ. Deformation mechanisms in crystalline polymers. J Mater Sci 1974;9(12):2034. 26. Kim JM, Locker R, Rutledge GC. Plastic deformation of semicrystalline polyethylene under extension, compression, and shear using molecular dynamics simulation. Macromolecules 2014;47(7):2515. 27. Lin L, Argon AS. Structure and plastic deformation of polyethylene. J Mater Sci 1994;29(2):294. 28. Takahashi Y, Shishido T, Yamamoto K, et al. Mechanisms of plastic deformation in highly cross-linked UHMWPE for total hip components e the molecular physics viewpoint. J Mech Behav Biomed Mater 2015;42:43. 29. Oral E, Christensen SD, Malhi AS, et al. Wear resistance and mechanical properties of highly cross-linked, ultrahigh-molecular weight polyethylene doped with vitamin E. J Arthroplasty 2006;21(4):580. 30. Wolf C, Krivec T, Blassnig J, et al. Examination of the suitability of alphatocopherol as a stabilizer for ultra-high molecular weight polyethylene used for articulating surfaces in joint endoprostheses. J Mater Sci Mater Med 2002;13(2):185. 31. Turner A, Okubo Y, Teramura S, et al. The antioxidant and non-antioxidant contributions of vitamin E in vitamin E blended ultra-high molecular weight polyethylene for total knee replacement. J Mech Behav Biomed Mater 2014;31:21. 32. Takahashi Y, Yamamoto K, Shishido T, et al. Strain-induced microstructural rearrangement in ultra-high molecular weight polyethylene for hip joints: a comparison between conventional and vitamin E-infused highly-crosslinked liners. J Mech Behav Biomed Mater 2014;31:31. 33. Muratoglu OK, Wannomae KK, Rowell SL, et al. Ex vivo stability loss of irradiated and melted UHMWPE. J Bone Joint Surg Am 2010;92(17):2809. 34. Oral E, Ghali BW, Neils AL, et al. A new mechanism of oxidation in UHMWPE caused by squalene absorption. J Biomed Mater Res B Appl Biomater 2012;100(3):742. 35. Beecher B, Huddleston J, Rowell S, et al. Material properties of a highly anteverted vitamin E-stabilized polyethylene liner after sixteen months in vivo. JBJS Case Connector 2014;4(3):e57. 36. Gigante A, Bottegoni C, Ragone V, et al. Effectiveness of vitamin-E-doped polyethylene in joint replacement: a literature review. J Funct Biomater 2015;6:889. 37. Scifert CF, Brown TD, Pedersen DR, et al. A finite element analysis of factors influencing total hip dislocation. Clin Orthop Relat Res 1998;355:152.