- Email: [email protected]
The biomaterials challenge: A comparison of polyethylene wear using a hip joint simulator
journal of the mechanical behavior of biomedical materials 53 (2016) 40 –48
Available online at www.sciencedirect.com
www.elsevier.com/locate/jmbbm
Research Paper
The biomaterials challenge: A comparison of polyethylene wear using a hip joint simulator Saverio Affatatoa,n, Nadia Frecceroa, Paola Taddeib a
Laboratorio di Tecnologia Medica, Istituto Ortopedico Rizzoli, Bologna, Italy Dipartimento di Scienze Biomediche e Neuromotorie, Università di Bologna, Via Belmeloro 8/2, 40126 Bologna, Italy
b
ar t ic l e in f o
abs tra ct
Article history:
Although hip arthroplasty is an established procedure that relieves pain and improves
Received 22 June 2015
functions, problems remain with wear and osteolysis. Highly cross-linked polyethylene
Received in revised form
and Vitamin-E-stabilized polyethylene were introduced in the last years to solve these
31 July 2015
problems.
Accepted 4 August 2015
In this study we compared the in vitro wear behaviour of cross-linked polyethylene
Available online 12 August 2015
(XLPE) versus Vitamin-E diffused XLPE (XLPE_VE) versus conventional ultra-high molecular
Keywords:
weight polyethylene (UHMWPE) acetabular cups. The test was performed using a hip joint
Vitamin-E doped polyethylene
simulator run for two millions cycles under bovine calf serum as lubricant.
Cross-linked polyethylene
Mass loss was found to decrease along the series UHMWPE4XLPE_VE4XLPE, although
Ultra-high molecular weight
statistically significant differences were found only between the mass losses of XLPE and
polyethylene
UHMWPE at 1.2 and 2 million cycles. The mass loss data were explained in relation to the
Hip simulator
crystalline morphology of the control unworn cups, as investigated by non-destructive
Raman spectroscopy
micro-Raman spectroscopy. This technique allowed to disclose a different wear behaviour of the three sets of cups. Wear testing produced a stress-induced crystallisation in UHMWPE, with increases in both amorphous (αa) and orthorhombic (αo) phases at the expense of the third phase (αb), which decreased upon wear. Moreover, the all-trans content decreased, while the ortho-trans content increased, contrarily to the trend observed for XLPE and XLPE_VE, for which no statistically significant changes in αo, αa and αb contents were detected. The XLPE_VE specimens underwent the least significant changes in the spectroscopic markers of micromorphology upon mechanical stress, probably due to their lower starting amorphous content. & 2015 Elsevier Ltd. All rights reserved.
1.
Introduction
failure (Brach del Prever et al., 2009; Bracco and Oral, 2011; James et al., 2009). Actually, during irradiation of UHMWPE,
It is well known that the oxidative degradation of ultra-high molecular weight polyethylene (UHMWPE) decreases its mechanical properties and represents the main cause of its n
Corresponding author. Tel.: þ39 051 6366864; fax: þ39 051 6366863. E-mail address: [email protected] (S. Affatato).
http://dx.doi.org/10.1016/j.jmbbm.2015.08.001 1751-6161/& 2015 Elsevier Ltd. All rights reserved.
free radicals are produced through the radiolytic cleavage of C–H and C–C bonds. They react with oxygen and trigger a complex cascade of reactions that lead to the formation of
journal of the mechanical behavior of biomedical materials 53 (2016) 40 –48
peroxy free radicals, hydroperoxides, and ultimately, carbonyl species, (mainly ketones, esters, and carboxylic acids). Thus, oxidation is accompanied by chain scission and reduction of the molecular weight of the polymer, possible recrystallisation, increase in stiffness and embrittlement, deteriorating the UHMWPE mechanical properties. The consequent increase in the wear debris production leads to biological responses that may cause osteolysis and implant loosening. The continuous and evolutive change from conventional UHMWPE to highly cross-linked polyethylenes has allowed many possible material combinations in the design of sliding couples for artificial hip joints. Highly cross-linked polyethylenes (XLPEs) were developed using high-dose irradiation (50– 100 kGy) to decrease the wear rate of UHMWPE (McKellop et al., 1999; Kurtz et al., 1999). Irradiation causes crosslinking in the amorphous phase of the UHMWPE, but also initiates the formation of free radicals in the crystalline phase; unable to recombine due to structural limitations, they become trapped for long periods of time (Oral et al., 2011; Singh 1999; Jahan et al., 2001; Micheli et al., 2012). These residual free radicals are believed to migrate to the crystalline/amorphous interface and cause oxidative degradation in the material through the above mentioned cascade of reactions with oxygen (Li et al., 2012; Bracco et al., 2007; Sobieraj and Rimnac, 2009; Mossman, 2008; Corcoran et al., 2009). The complete elimination of free radicals is difficult due to the limited mobility of these radicals within the crystalline structure of the polymer. Post-irradiation re-melting has been proposed to allow the recombination of free radicals trapped within the crystalline regions, with a consequent improvement in the oxidation and wear resistance (McKellop et al., 1999); unfortunately, remelting results in a decrease of PE crystallinity and thus in a loss of strength (Oral et al., 2006). Another strategy to retard oxidation consists in the addition of suitable stabilising additives, whose role is to decrease the reactivity of the radical species, to slow the oxidation processes and to preserve the PE chemical, physical and mechanical properties, i.e. to prolong its lifetime (Zweifel, 1998; Kurtz et al., 1999). Vitamin E seems to be the ideal candidate, being already employed as a natural antioxidant in the physiological processes of the human body (Bracco et al., 2007). The rationale for using this additive was the improvement of the oxidation resistance of irradiated UHMWPE as well as its fatigue strength, using an alternative to post-irradiation remelting (Bracco and Oral, 2011; Oral et al., 2008, 2010). Two methods may be used to incorporate vitamin E into PE, i.e. before consolidation and crosslinking or after irradiation. Besides prevention of oxidative degradation, the incorporation of vitamin E into the microstructure of PE components has been reported to determine a reduced incidence of fatigue crack, a reduced biological response and osteolytic potential induced by wear particle (Azzi and Stocker, 2000; Oral and Muratoglu, 2011). It is unclear if antioxidants improve UHMWPE durability in vivo by reducing oxidation and show less abrasion in the long-term. To date, there are no long-term clinical data available comparing XLPE with vitamin E stabilized UHMWPE/XLPE in hip arthroplasty liners; an in vivo evaluation of PE oxidation is not possible
41
(Kurtz et al., 2009; Lerf et al., 2010). At this moment, a longterm in vivo randomised controlled trial has been only designed to determine wear, bone mineral density, functional outcome and survival (Van der Veen et al., 2012). Salemyr et al. (2015) have carried out a prospective, randomized, controlled trial aimed at comparing a new vitamin-E diffused XLPE liner and a conventional XLPE liner, at a two-years followup; they have reported a lower initial head penetration and lower superior and medial wear for the former. However, they have acknowledged that a longer followup is necessary to demonstrate lower overall head penetration. In laboratory-based studies, various attempts have been made to compare the wear behaviour of conventional PE and vitamin-E stabilised liners (Affatato et al., 2010, 2012). However, the application of these models to the clinical situation did not appear satisfactory. Discrepancies have been observed on the role of the vitamin E between the results of laboratory wear tests and clinical observations. Moreover, contradictory results have been reported for in vitro studies using joint simulators. Affatato et al. (2012) in in vitro wear tests found that the vitamin-E blended XLPE wore more than XLPE and conventional UHMWPE. Previous studies have shown a reduction in wear rate with the addition of vitamin E to highly cross-linked UHMWPE compared to conventional UHMWPE, but in many of them, it is not clear whether the reduction in wear rate was due to the addition of vitamin E or the high levels of cross-linking (Micheli et al., 2012; Oral et al., 2006; Wannomae et al., 2010). Bladen et al. (2013) have reported that the addition of vitamin E to nonirradiated UHMWPE did not affect the wear factor suggesting that in previous studies, the reduction in wear was more likely to be associated with the high levels of crosslinking. On the other hand, Teramura et al. (2009) have reported that the wear volume of vitamin E-blended UHMWPE tested with a knee joint simulator was 30% lower than that of virgin UHMWPE after 5 million cycles (Mc). In this context, the main goal of this study was to comparatively investigate the wear performance of vitamin E-stabilised XLPE, XLPE and conventional UHMWPE acetabular cups in a hip joint simulator study. To gain more insights into the role of vitamin E in the wear mechanism of acetabular cups, the samples were analysed by micro-Raman spectroscopy. This technique proved suitable to investigate on a molecular scale the morphology changes (i.e. in crystallinity, phases distribution, chain orientation, residual strain) induced by wear. Micro-Raman spectroscopy was preferred over other techniques of molecular characterisation due to its nondestructive character, in view of continuing the tests under more severe conditions (i.e. after accelerated ageing and in the presence of third-body wear particles).
2.
Materials and methods
The wear behaviour of three different batches of polyethylene acetabular cups (32-mm inner 50-mm outer dimensions; 5 specimens for each batch) coupled with 32-mm cobalt–chromium–molybdenum (CoCrMo) femoral heads was investigated using a hip joint simulator. In particular, four components of each batch run onto the simulator and,
42
journal of the mechanical behavior of biomedical materials 53 (2016) 40 –48
following a standardized procedure (Affatato et al., 2005), another three acetabular cups (one for each type of material used) were stored (non-loaded) in bovine calf serum to compensate for weight changes due to fluid absorption. All polyethylenes tested in this study were machined from polymer bars Chirulen GUR 1020 (Polymax, Adler, Milan, Italy). Cross-linked acetabular cups (XLPE) were firstly γ-ray irradiated to 75 kGy (710%), then thermally treated at 150 1C, in order to remove free radicals formed during irradiation. After these treatments, the cups were machined to their final shape. Similarly, Vitamin E-containing (0.1%), XLPE acetabular cups (hereinafter called XLPE_VE) were machined from a Vitamin E-blended UHMWPE bars (Polymax, Adler, Milan, Italy), after electron beam irradiation to 75 kGy (710%), followed by a thermal treatment at 150 1C under nitrogen for 12 h. All polyethylene acetabular cups were pre-soaked for four weeks prior the wear tests.
2.1.
Wear test details
Wear tests were performed using a new 12-station hip joint simulator (IORSynthe, Bologna, Italy). The improved software of this hip simulator allows applying the kinematics as per ISO 14242-1:2012. The test was carried out after validation of the hip joint simulator; the ability of the hip simulator to follow the kinematic inputs and outputs as recommended by ISO 14242-1:2012 was verified. With the versatility of the improved software, all degree of freedom (DOF) was monitored, aligned, and modified in order to obtain all kinematic inputs fixed by the international recommendations and also to validate wear rates between individual stations. After validation, the UHMWPE, XLPE and XLPE_VE acetabular cups (three specimens for each set) were tested for two Mc. The simulator utilised hydraulic actuators to apply the cyclic vertical compressive loads (oscillating between 300 and 3000 N applied perpendicular to the acetabular components as recommend by the International ISO 14242). The lubricant used was 25 vol% new born calf serum balanced with distilled water, with 0.2% sodium azide in order to retard bacterial growth and 20 mM EDTA (ethylenediaminetetracetic acid) to minimise precipitation of calcium phosphate. The mass loss of the cups was determined every 0.4 million cycles using a new microbalance (SARTORIUS CUBIS MSE 225S-000-DU, Goettingën, Germany) with a sensitivity of 0.01 mg and an uncertainty of 0.01 mg. Each mass measurement was repeated three times and the average mass was used for calculations. Wear trend was determined from the mass loss of each acetabular cup, corrected by acetabular soak control; the wear rates, calculated from the steady-state slopes of the mass loss versus number of cycles lines, were obtained using least squares linear regression. The mass loss data were analysed using a nonparametric Kruskall–Wallis (K–W) test and a least significance difference as post-hoc test. Statistical significance was set at po0.05.
2.2.
Micro-Raman analyses
The PE components that showed the highest mass losses (two UHMWPE, one XLPE and one XLPE_VE) were analysed by micro-Raman spectroscopy since they were expected to show
the most significant changes on a molecular scale. The unworn control samples were characterised as well. MicroRaman spectra were obtained using a Jasco NRS-2000C instrument with a microscope of 50 magnification. All the spectra were recorded in back-scattering conditions with 5 cm 1 spectral resolution using the 532 nm Green Diode Pumped Solid State (DPSS) Laser Driver (RgBLase LLC, USA) with a power of ca. 20 mW. A 160 K cooled digital CCD (Spec-10: 100B, Roper Scientific Inc.) was used as a detector. A confocal pinhole with an aperture diameter of 200 μm was placed in the optical circuit to obtain signals from a limited in-depth region. All the Raman measurements were performed in a fully non-destructive way, without any sample manipulation. The worn components were analysed near the centre of the articulating surface, which appeared the most damaged area according to previously reported data (Bertoluzza et al., 2000; Affatato et al., 2003); twelve spectra at least were recorded. The same number of spectra was recorded on the unworn components. At a first approximation, PE can be considered as composed by three different phases (Strobl and Hagedorn, 1978): a crystalline phase (with the crystals being primarily orthorhombic in structure (Taddei et al., 2008, 2011)), a melt-like amorphous phase and an anisotropic disordered phase (i.e. a “third phase” with a prevailing trans conformation of the chains). Analogously to a previous study (Tozzi et al., 2014), we evaluated the effects of wear testing on the PE structure through several spectroscopic markers proposed in the literature (Strobl and Hagedorn, 1978; Lagaron et al., 1999). The fraction of orthorhombic (αo), amorphous (αa) and intermediate anisotropic disordered (αb) phases was calculated from the relative intensities of selected Raman bands, according to Strobl and Hagedorn (1978): αo ¼
A1414 0:46 A1295þ1305
ð1Þ
αa ¼
A1080 0:79 A1295þ1305
ð2Þ
αb ¼ 1 ðαo þ αa Þ
ð3Þ
where A1414 and A1080 are the areas of the Raman bands at 1414 and 1080 cm 1, respectively; A1295þ1305 is the area of the internal standard band group. The A1080 band area was determined trough curve fitting analysis by means of a commercial software (Opus 5.0 from Bruker Optik GmbH, Germany). Prior to curve fitting, the original spectra were baseline corrected. The Raman components were described as linear combinations of Gaussian and Lorentzian functions and curve fitting was performed using the LevenbergMarquardt algorithm. According to previous studies (Taddei et al., 2011; Kurelec et al., 2000), the occurrence of an orthorhombic to monoclinic phase transformation was evaluated through the I1414/ (I1440þI1460) intensity ratio, where I1414, I1440 and I1460 were calculated as peak heights. According to Lagaron et al. (1999), the fraction of all-trans sequences was evaluated through the following ratio: all trans ¼
A1130 0:80 A1295þ1305
ð4Þ
journal of the mechanical behavior of biomedical materials 53 (2016) 40 –48
where A1130 is the area of the Raman band at about 1130 cm 1, which has been associated with all-trans C–C bonds located both in the crystalline and amorphous phases. The fraction of all-trans sequences in an orthorhombic environment (i.e. ortho-trans) was calculated as ortho trans ¼
1:78 A1414 A1130
ð5Þ
The occurrence of orientation upon wear tests was assessed trough the A1130/A1060 area ratio. The 1060 cm 1 band area was calculated by curve fitting, according to the same procedure above reported. The residual strain of polyethylene was evaluated through the full-width at halfmaximum (FWHM) of the bands at about 1130 and 1060 cm 1.
3.
Results
All the prosthetic hip specimens completed the planned two Mc. As shown in Fig. 1, the XLPE_VE acetabular cups wore more than the XLPE ones, although the results of the posthoc test (Table 1) did not disclose any significant difference between the two sets of cups at any number of cycles. The XLPE_VE acetabular cups maintained a lower mass loss (although not statistically significant, Table 1) than the UHMWPE ones during the whole test. A statistically significant difference (p¼ 0.034) was observed between the XLPE and UHMWPE acetabular cups run onto the simulator for 1.2 and 2 million cycles (Table 1). No statistically significant differences were observed between the three sets of polyethylene cups using the K–S statistical test at the other number of cycles. Fig. 2 reports the micro-Raman spectra (normalised to the A1295þ1305 area) of the control unworn UHMWPE, XLPE and XLPE_VE acetabular cups, in the CH2 bending, CH2 twisting and C–C stretching regions; the assignments to the different phases have been given according to the literature (Mutter et al., 1993). As can be easily seen, they are not perfectly coincident, especially with regards to the bands at about 1440 and 1460 cm 1 (CH2 bending region) and 1160 cm 1 (C–C stretching region). The band at 1414 cm 1, assignable to the orthorhombic crystalline structure, showed nearly the same intensity in the
43
three spectra; accordingly, the αo orthorhombic content in the three sets of acetabular cups was nearly the same and no significantly differences were detected (p ¼0.056, Fig. 4A). The band at 1080 cm 1, assignable to the amorphous phase, appeared slightly weaker in the spectrum of the XLPE_VE control cup (Fig. 2); actually, the statistical analysis showed for this specimen a significantly lower αa amorphous content value than for UHMWPE and XLPE (UHMWPE versus XLPE_VE: po0.001; XLPE versus XLPE_VE: p ¼0.001, Fig. 3A). No significant differences were detected between the αb third phase content values of the three sets of cups (p ¼0.54, Fig. 3A). Fig. 3B shows statistically significant differences between UHMWPE and XLPE (in all-trans and ortho-trans contents, A1130/A1060; po0.001) as well as between UHMWPE and XLPE_VE (in all-trans and ortho-trans contents, A1130/ A1060 (po0.001) and I1414/(I1440þI1460) p ¼0.045). From the data reported in Fig. 3A and B, it can be easily seen that the spectroscopic markers corresponding to XLPE and XLPE_VE control acetabular cups had an associated standard deviation higher than for UHMWPE. For most spectroscopic markers, the standard deviation associated to XLPE_VE was the highest. The differences in the all-trans and ortho-trans contents as well as in the A1130/A1060 ratio (Fig. 3B) reflect the differences in intensity of the band at about 1130 cm 1, which is significantly higher in UHMWPE than in XLPE and XLPE_VE (Fig. 2). The I1414/(I1440þI1460) ratio is significantly higher in XLPE_VE (Fig. 3B), as expectable on the basis of the significantly lower intensity of the bands at 1440 and 1460 cm 1, due to the higher amorphous content present in this sample, also revealed by the higher αa value (Fig. 3A). The three sets of unworn samples were characterised by significant differences in the FWHM values attained by both the 1130 and 1060 cm 1 bands (Fig. 3C), suggesting that crosslinking and addition of vitamin E modified the residual strain in the material. The data reported in Figs. 4 and 5 showed the effects of wear on the micromorphology of the three sets of cups, as investigated through the αo, αa, αb, all-trans and ortho-trans contents, I1414/(I1440þI1460) and A1130/A1060 ratios as well as the
Fig. 1 – Wear behaviour and regression coefficient for the different polyethylenes tested.
44
journal of the mechanical behavior of biomedical materials 53 (2016) 40 –48
Table 1 – Cumulative mass loss (mean7standard deviation) for the three sets of PE acetabular cups tested. Values and statistical analysis performed using a Kruskall–Wallis nonparametric test. Mean7standard deviation
K–W test (p-value)
Cycles [Mc]
UHMWPE
XLPE
XLPE_VE
0.4 0.8 1.2 1.6 2
5.270.5 13.170.6 24.571.7 27.971.9 35.473.3
1.870.7 2.771.2 3.371.5 4.772.7 6.774.0
1.871.0 3.871.5 8.975.5 11.577.5 16.178.2
0.066 0.066 0.039 0.393 0.039
Post-hoc test (p-value) UHMWPE versus XLPE
UHMWPE versus XLPE_VE
XLPE versus XLPE_VE
0.54 0.33 0.034 0.33 0.034
0.54 0.76 0.76 0.76 0.76
1.00 0.76 0.76 0.76 0.76
NS ¼not significant (p value40.05); K–W ¼Kruskall–Wallis; Mc ¼millions cycles.
Fig. 2 – Micro-Raman spectra (normalised to the A1295þ1305 area) of the control unworn UHMWPE (black), XLPE (blue) and XLPE_VE (red) acetabular cups, in the CH2 bending, CH2 twisting and C–C stretching regions; the assignments to the amorphous (A) and crystalline (C) phases have been given according to the literature. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
FWHM1130 and FWHM1060. Upon wear, the UHMWPE acetabular cups showed a significant increase of αo (p¼ 0.041 for #2) αa (p ¼0.010 and o0.001 for #1 and #2, respectively, Fig. 4) and ortho-trans contents (po0.001 and p ¼0.001 for #1 and #2, respectively, Fig. 5) and a significant decrease of αb (p¼ 0.001 for #2, Fig. 4) and all-trans contents (po0.001 and p ¼0.04 for #1 and #2, respectively, Fig. 5), I1414/(I1440þI1460) (p ¼0.008 for #1, Fig. 5) and A1130/A1060 ratios (p¼ 0.045 and 0.012 for #1 and #2, respectively, Fig. 5). More significant changes in the FWHM1130 and FWHM1060 were observed for the sample #2, suggesting that wear has modified the residual strain in this component (Fig. 5). If compared with UHMWPE, the XLPE and XLPE_VE acetabular cups showed a different trend of the spectroscopic
markers upon wear and appeared similar each other: the alltrans content significantly increased (p¼ 0.004 and po0.001 for XLPE and XLPE_VE, respectively, Fig. 5) as well as the A1130/A1060 ratio (po0.001 for both XLPE and XLPE_VE, Fig. 5), while the ortho-trans content decreased (p ¼0.0017 and p ¼0.004 for XLPE and XLPE_VE, respectively, Fig. 5). No significant changes were observed for αo, αa and αb contents.
4.
Discussion
Due to contradicting results on the in vitro wear performance of vitamin E-containing PE (Affatato et al., 2010, 2012; Wannomae et al., 2010; Bladen et al., 2013; Teramura et al., 2009), a hip joint
journal of the mechanical behavior of biomedical materials 53 (2016) 40 –48
45
Fig. 3 – Values (average7standard deviation) of αo, αa and αb contents (A), all-trans and ortho-trans contents, I1414/ (I1440þI1460) and A1130/A1060 ratios (B), FWHM1130 and FWHM1060 (C), as obtained from the spectra of the control unworn UHMWPE, XLPE and XLPE_VE acetabular cups. Arrows with asterisks indicate statistically significant differences (po0.05).
Fig. 4 – Values (average7standard deviation) of αo, αa and αb contents, as calculated from the spectra of the control unworn and worn UHMWPE, XLPE and XLPE_VE acetabular cups. Arrows with asterisks indicate statistically significant differences with respect to control cups (po0.05).
simulator study has been carried out. Using a 12-station hip simulator, the wear performance of three different sets of PE acetabular cups coupled with CoCrMo femoral heads was evaluated for two million cycles; we asked whether the addition of vitamin E to XLPE would improve the wear behaviour of such material with respect to the conventional UHMWPE and XLPE. The results of this investigation clearly showed a reduced wear for the XLPE and the XLPE_VE acetabular cups with respect to the conventional UHMWPE components. However, a statistically significant difference was observed only between the XLPE and UHMWPE acetabular cups run onto the simulator for 1.2 and 2 million cycles (Table 1). No statistically significant differences were observed between the three sets of polyethylene cups using the K–S statistical test at the other number of cycles.
To gain insights into these results on a molecular scale, micro-Raman spectroscopic measurements on control and simulator tested UHMWPE, XLPE and XLPE_VE acetabular cups were carried out. With regards to the control cups, the XLPE_VE unworn cup showed a statistically significant lower αa amorphous content (Fig. 3A) than the other unworn cups. With regards to the all-trans and orthotrans contents as well as the A1130/A1060 ratio, XLPE and XLPE_VE control cups did not appear significantly different each other, while showed statistically significant differences with UHMWPE (Fig. 3B). On the basis of the obtained results, it may be affirmed that crosslinking and post-irradiation melting did not sensibly alter the distribution of the crystalline and amorphous phases (i.e. orthorhombic, amorphous and third intermediate phase),
46
journal of the mechanical behavior of biomedical materials 53 (2016) 40 –48
Fig. 5 – Values (average7standard deviation) of all-trans and ortho-trans contents, I1414/(I1440þI1460) and A1130/A1060 ratios, FWHM1130 and FWHM1060, as calculated from the spectra of the control unworn and worn UHMWPE, XLPE and XLPE_VE acetabular cups. Arrows with asterisks indicate statistically significant differences with respect to control cups (po0.05). since their content in XLPE control cup was nearly the same as in UHMWPE (Fig. 3A). However, these processes induced structural rearrangements that affected the all-trans and ortho-trans contents (the former decreased, the latter increased, Fig. 3B), as well as the chain orientation (as revealed by the decrease of the A1130/A1060 ratio, Fig. 3B). Actually, the bands at 1130 and 1060 cm 1 have different vibrational symmetries; if the molecules are oriented in a preferred direction, the 1130 cm 1 band has been reported to become stronger with respect to the 1060 cm 1 band (Pigeon et al., 1991). The presence of vitamin E during cross-linking and postirradiation did not sensibly affect the value of the latter spectroscopic markers, while determined a decrease of the amorphous phase content. The differences observed in the control acetabular cups may have determined the differences in the wear behaviour
of the three sets of acetabular cups. Actually, it is well recognised that the mechanical properties of PE depend on the distribution of crystalline and non-crystalline phases as well as molecular orientation, besides molecular weight and degree of cross-linking. The latter aspect may explain the trend of the mass loss which was found to decrease along the series UHMWPE4XLPE_VE4XLPE, although significant differences were observed only between UHMWPE and XLPE at 1.2 and 2 Mc. It is well recognised that under multidirectional stress conditions an anisotropic PE surface leads to a decreased resistance to adhesive wear due to strain-softening and weakening. These adverse effects are mitigated by cross-linking, due to the increased number of intralamellar and inter-lamellar covalent bonds. Therefore, the results of the present test confirmed the positive effects of cross-linking on wear resistance.
journal of the mechanical behavior of biomedical materials 53 (2016) 40 –48
The morphological differences observed by micro-Raman spectroscopy in the control acetabular cups determined different wear behaviours at molecular level (i.e. different trends of the spectroscopic markers upon wear). As stressed above, the main spectroscopic differences were observed by comparing control UHMWPE versus XLPE and XLPE_VE (Fig. 3); actually, XLPE and XLPE_VE appeared quite similar each other. As expectable, they underwent similar changes upon wear, which appeared significantly different (i.e. opposite in several aspects) to those experienced by UHMWPE (Figs. 4 and 5). Wear testing proved to induce a stress-induced crystallisation in UHMWPE, in agreement with previous studies (Affatato et al., 2005, 2013); actually, the αo content appeared to significantly increase upon wear (Fig. 4), as a result of the near-surface orientation of the crystalline lamellae observed in the wear-tested acetabular liners (Edidin et al., 1999). The αo increase upon wear in XLPE and XLPE_VE did not appear significant, in agreement with previous studies (Affatato et al., 2005); actually, the thickness of the plasticity-induced damage layer that precedes abrasive wear has been reported to be lower in XLPE. Also the αa amorphous content appeared to significantly increase upon wear in UHMWPE (Fig. 4); both αa and αo increased at the expense of αb, which decreased upon wear. At the same time, the all-trans content decreased while the ortho-trans content increased (Fig. 5). XLPE and XLPE_VE showed an opposite trend upon wear: actually, the ortho-trans content decreased, while the all-trans content increased (Fig. 5); no statistically significant changes in αo, αa and αb were detected. With regards to XLPE_VE, it must also be observed that vitamin E has been suggested to be present within the amorphous phase of PE (Mallégol et al., 2001) and for its high affinity for the polymer, it inhibits the strain-induced crystallisation phenomenon (Okubo et al., 2009). Also molecular orientation effects appeared different in UHMWPE versus XLPE and XLPE_VE. In the former, the A1130/A1060 ratio decreased upon wear (Fig. 5), while increased in the crosslinked specimens (i.e. XLPE and XLPE_VE). The spectroscopic results disclosed a different wear mechanism between UHMWPE and cross-linked PEs. It must be stressed that upon wear, the XLPE_VE acetabular cups appeared the least affected by mechanical stress; actually, this kind of specimens underwent the least significant changes of the various spectroscopic markers investigated (Figs. 4 and 5). Probably, these slighter morphological alterations induced by wear may be related to its lower starting αa content (Fig. 3A). Actually, a lower content of the amorphous phase has been related to a higher resistance to creep deformation, but to a lower capacity of shape recovery after removal of the applied load (Takahashi et al., 2014).
5.
Conclusions
The acetabular cups studied in this work showed a mass loss upon wear testing decreasing along the series UHMWPE4XLPE_VE4XLPE, although statistically significant differences were found only between the mass losses of XLPE and UHMWPE at 1.2 and 2 million cycles. The mass loss data were explained in relation to molecular morphology, as
47
investigated by non-destructive micro-Raman spectroscopy. This technique allowed to disclose a different wear behaviour of the UHMWPE and cross-linked PE specimens (i.e. an opposite trend of several spectroscopic markers); the vitamin E-blended specimens underwent the least significant micromorphological changes upon mechanical stress.
Acknowledgements The authors would like to thank Sami Abdel Jaber for his help during the set-up and Tommaso Vannini for his help with statistical analyses. Thanks are due also to Riccardo Verga (Research & Development, Adler Ortho S.r.l, Milan, Italy) for his supplying of the samples. The authors would like to express their gratitude to the Synthesis (Florence, Italy) for using the simulator. This work was partially supported by the Italian Program of Donation for Research “5 per mille”, (Cardinis 5136) year 2011.
r e f e r e nc e s
Affatato, S., Modena, E., Carmignato, S., Taddei, P., 2013. The use of Raman spectroscopy in the analysis of UHMWPE unicondylar bearing systems after run on a force and displacement control knee simulators. Wear 297, 781–790. Affatato, S., Bersaglia, G., Emiliani, D., Foltran, I., Taddei, P., Reggiani, M., et al., 2003. The performance of gamma- and EtO-sterilised UHMWPE acetabular cups tested under severe simulator conditions. Part 2: wear particle characteristics with isolation protocols. Biomaterials 24, 4045–4055. Affatato, S., Bersaglia, G., Rocchi, M., Taddei, P., Fagnano, C., Toni, A., 2005. Wear behaviour of cross-linked polyethylene assessed in vitro under severe conditions. Biomaterials 26, 3259–3267. Affatato, S., Zavalloni, M., Spinelli, M., Costa, L., Bracco, P., Viceconti, M., 2010. Long-term in-vitro wear performance of an innovative thermo-compressed cross-linked polyethylene. Tribol. Int. 43, 22–28. Affatato, S., Bracco, P., Costa, L., Villa, T., Quaglini, V., Toni, A., 2012. In vitro wear performance of standard, crosslinked, and vitamin-E-blended UHMWPE. J. Biomed. Mater. Res. A 100, 554–560. Azzi, A., Stocker, A., 2000. Vitamin E: non-antioxidant roles. Prog. Lipid Res. 39, 231–255. Bertoluzza, A., Fagnano, C., Rossi, M., Tinti, A., Cacciari, G.L., 2000. Micro-Raman spectroscopy for the crystallinity characterization of UHMWPE hip cups run on joint simulators. J. Mol. Struct. 521, 89–95. Bladen, C.L., Teramura, S., Russell, S.L., Fujiwara, K., Fisher, J., Ingham, E., et al., 2013. Analysis of wear, wear particles, and reduced inflammatory potential of vitamin E ultrahighmolecular-weight polyethylene for use in total joint replacement. J. Biomed. Mater. Res. B 101 B, 458–466. Bracco, P., Oral, E., 2011. Vitamin E-stabilized UHMWPE for total joint implants: a review. Clin. Orthop. Relat. Res. 469, 2286–2293. Bracco, P., Brunella, V., Zanetti, M., Luda, M.P., Costa, L., 2007. Stabilisation of ultra-high molecular weight polyethylene with vitamin E. Polym. Degrad. Stab. 92, 2155–2162. Brach del Prever, E.M., Bistolfi, A., Bracco, P., Costa, L., 2009. UHMWPE for arthroplasty: past or future?. J. Orthop. Traumatol. 10, 1–8.
48
journal of the mechanical behavior of biomedical materials 53 (2016) 40 –48
Corcoran, P.L., Biesinger, M.C., Grifi, M., 2009. Plastics and beaches: a degrading relationship. Mar. Pollut. Bull. 58, 80–84. Edidin, A.A., Pruitt, L., Jewett, C.W., Crane, D.J., Roberts, D., Kurtz, S.M., 1999. Plasticity-induced damage layer is a precursor to wear in radiation-cross-linked UHMWPE acetabular components for total hip replacement. J. Arthroplast. 14, 616–627. Jahan, M.S., King, M.C., Haggard, W.O., Sevo, K.L., Parr, J.E., 2001. A study of long-lived free radicals in gamma-irradiated medical grade polyethylene. Radiat. Phys. Chem. 62, 141–144. James S.P., Oldinski R. , Zhang M., Schwartz H., 2009. UHMWPE Biomaterials Handbook. Kurelec, L., Rastogi, S., Meier, R.J., Lemstra, P.J., 2000. Chain mobility in polymer systems: on the borderline between solid and melt. 3. Phase transformations in nascent ultra high molecular weight polyethylene reactor powder at elevated pressure as revealed by in-situ Raman spectroscopy. Macromolecules 33, 5593–5601. Kurtz, S.M., Muratoglu, O.K., Evans, M., Edidin, A.A., 1999. Advances in the processing, sterilization, and crosslinking of ultra-high molecular weight polyethylene for total joint arthroplasty. Biomaterials 20, 1659–1688. Kurtz, S.M., Dumbleton, J., Siskey, R.S., Wang, A., Manley, M., 2009. Trace concentrations of vitamin E protect radiation crosslinked UHMWPE from oxidative degradation. J. Biomed. Mater. Res. A 90, 549–563. Lagaron, J.M., Dixon, N.M., Reed, W., Pastor, J.M., Kip, B.J., 1999. Morphological characterisation of the crystalline structure of cold-drawn HDPE used as a model material for the environmental stress cracking (ESC) phenomenon. Polymer 40, 2569–2586. Lerf, R., Zurbru¨gg, D., Delfosse, D., 2010. Use of vitamin E to protect cross-linked UHMWPE from oxidation. Biomaterials 31, 3643–3648. Li, C.S., Zhan, M.S., Huang, X.C., Zhou, H., 2012. Degradation behavior of ultra-high molecular weight polyethylene fibers under artificial accelerated weathering. Polym. Test. 31, 938–943. Malle´gol, J., Carlsson, D.J., Descheˆnes, L., 2001. Post-γ-irradiation reactions in vitamin E stabilised and unstabilised HDPE. Nucl. Instrum. Methods Phys. Res. B 185, 283–293. McKellop, H., Shen, F.W., Lu, B., Campbell, P., Salovey, R., 1999. Development of an extremely wear-resistant ultra high molecular weight polyethylene for total hip replacements. J. Orthop. Res. 17, 157–167. Micheli, B.R., Wannomae, K.K., Lozynsky, A.J., Christensen, S.D., Muratoglu, O.K., 2012. Knee simulator wear of vitamin E stabilized irradiated ultrahigh molecular weight polyethylene. J. Arthroplast. 27, 95–104. Mossman, S., 2008. Conservation of plastics: materials science, degradation and preservation. Nature 455, 288–289. Mutter, R., Stille, W., Strobl, G., 1993. Transition regions and surface melting in partially crystalline polyethylene: a Raman spectroscopic study. J. Polym. Sci. B: Polym. Phys. 31, 99–105. Okubo, Y., Teramura, S., Niwa, Y., Ibaraki, K., Murata, K., Hyon, S. H., et al., 2009. Strain-induced crystallization and orientation of vitamin E-blended ultrahigh molecular weight polyethylene. Biomed. Mater. Eng. 19, 431–439. Oral, E., Muratoglu, O.K., 2011. Vitamin E diffused, highly crosslinked UHMWPE: a review. Int. Orthop. 35, 215–223. Oral, E., Christensen, S.D., Malhi, A.S., Wannomae, K.K., Muratoglu, O.K., 2006. Wear resistance and mechanical properties of highly cross-linked, ultrahigh-molecular weight polyethylene doped with vitamin E. J. Arthroplast. 21, 580–591. Oral, E., Malhi, A.S., Muratoglu, O.K., 2006. Mechanisms of decrease in fatigue crack propagation resistance in irradiated and melted UHMWPE. Biomaterials 27, 917–925.
Oral, E., Malhi, A.S., Wannomae, K.K., Muratoglu, O.K., 2008. Highly cross-linked ultrahigh molecular weight polyethylene with improved fatigue resistance for total joint arthroplasty. Recipient of the 2006 Hap Paul Award. J. Arthroplast. 23, 1037–1044. Oral, E., Ghali, B.W., Rowell, S.L., Micheli, B.R., Lozynsky, A.J., Muratoglu, O.K., 2010. A surface crosslinked UHMWPE stabilized by vitamin E with low wear and high fatigue strength. Biomaterials 31, 7051–7060. Oral, E., Ghali, B.W., Muratoglu, O.K., 2011. The elimination of free radicals in irradiated UHMWPEs with and without vitamin e stabilization by annealing under pressure. J. Biomed. Mater. Res. B 97 B, 167–174. Pigeon, M., Prud’homme, R.E., Pezolet, M., 1991. Characterization of molecular-orientaiton in polyethyleen by Ramanspectroscopy. Macromolecules 24, 5687–5694. Salemyr, M., Muren, O., Ahl, T., Bode´n, H., Chammout, G., Stark, A., et al., 2015. Vitamin-E diffused highly cross-linked polyethylene liner compared to standard liners in total hip arthroplasty. A randomized, controlled trial. Int. Orthop. 39 (8), 1499–1505. Singh, A., 1999. Irradiation of polyethylene: some aspects of crosslinking and oxidative degradation. Radiat. Phys. Chem. 56, 375–380. Sobieraj, M.C., Rimnac, C.M., 2009. Ultra high molecular weight polyethylene: mechanics, morphology, and clinical behavior. J. Mech. Behav. Biomed. Mater. 2, 433–443. Strobl, G.R., Hagedorn, W., 1978. Raman spectroscopic method for determining the crystallinity of polyethylene. J. Polym. Sci.: Polym. Phys. Ed. 16, 1181–1193. Taddei, P., Affatato, S., Rocchi, M., Fagnano, C., Viceconti, M., 2008. The effects of irradiation and EtO-treatment on ultrahigh molecular weight polyethylene acetabular cups following accelerated aging: degradation of mechanical properties and morphology changes during hip simulator tests. J. Mol. Struct. 875, 254–263. Taddei, P., Di Foggia, M., Affatato, S., 2011. Raman characterization of conventional and cross-linked polyethylene in acetabular cups run on a hip joint simulator. J. Raman Spectrosc. 42, 1344–1352. Takahashi Y., Masaoka T., Pezzotti G., Shishido T., Tateiwa T., Kubo K., et al., 2014. Highly cross-linked polyethylene in total hip and knee replacement : spatial distribution of molecular orientation and shape recovery behavior, BioMed. Research International, Vol. 2014 (2014), 1-13, http://dx.doi.org/10.1155/ 2014/808369. Teramura, S., Sakoda, H., Terao, T., Fujiwara, K., Kawai, K., Tomita, N., 2009. Reduction of wear volume from accelerated aged UHMWPE knee components by the addition of vitamin E. J. Biomech. Sci. Eng. 4, 589–596. Tozzi, S., Modena, E., Falcioni, S., Sudanese, A., Affatato, S., Taddei, P., 2014. The effects of contact area and applied load on the morphology of in vitro worn ultra-high molecular weight knee prostheses: a micro-Raman and gravimetric study. J. Raman Spectrosc. 45, 781–787. Van der Veen, H.C., van den Akker-Scheek, I., Bulstra, S.K., van Raay, J.J., 2012. Wear, bone density, functional outcome and survival in vitamin E-incorporated polyethylene cups in reversed hybrid total hip arthroplasty: design of a randomized controlled trial. BMC Musculoskelet. Disord. 13, 178. Wannomae, K.K., Christensen, S.D., Micheli, B.R., Rowell, S.L., Schroeder, D.W., Muratoglu, O.K., 2010. Delamination and adhesive wear behavior of α-tocopherol-stabilized irradiated ultrahigh-molecular-weight polyethylene. J. Arthroplast. 25, 635–643. Zweifel, H., 1998. Stabilization of Polymeric Materials. Springer, Berlin, Germany.
Available online at www.sciencedirect.com
www.elsevier.com/locate/jmbbm
Research Paper
The biomaterials challenge: A comparison of polyethylene wear using a hip joint simulator Saverio Affatatoa,n, Nadia Frecceroa, Paola Taddeib a
Laboratorio di Tecnologia Medica, Istituto Ortopedico Rizzoli, Bologna, Italy Dipartimento di Scienze Biomediche e Neuromotorie, Università di Bologna, Via Belmeloro 8/2, 40126 Bologna, Italy
b
ar t ic l e in f o
abs tra ct
Article history:
Although hip arthroplasty is an established procedure that relieves pain and improves
Received 22 June 2015
functions, problems remain with wear and osteolysis. Highly cross-linked polyethylene
Received in revised form
and Vitamin-E-stabilized polyethylene were introduced in the last years to solve these
31 July 2015
problems.
Accepted 4 August 2015
In this study we compared the in vitro wear behaviour of cross-linked polyethylene
Available online 12 August 2015
(XLPE) versus Vitamin-E diffused XLPE (XLPE_VE) versus conventional ultra-high molecular
Keywords:
weight polyethylene (UHMWPE) acetabular cups. The test was performed using a hip joint
Vitamin-E doped polyethylene
simulator run for two millions cycles under bovine calf serum as lubricant.
Cross-linked polyethylene
Mass loss was found to decrease along the series UHMWPE4XLPE_VE4XLPE, although
Ultra-high molecular weight
statistically significant differences were found only between the mass losses of XLPE and
polyethylene
UHMWPE at 1.2 and 2 million cycles. The mass loss data were explained in relation to the
Hip simulator
crystalline morphology of the control unworn cups, as investigated by non-destructive
Raman spectroscopy
micro-Raman spectroscopy. This technique allowed to disclose a different wear behaviour of the three sets of cups. Wear testing produced a stress-induced crystallisation in UHMWPE, with increases in both amorphous (αa) and orthorhombic (αo) phases at the expense of the third phase (αb), which decreased upon wear. Moreover, the all-trans content decreased, while the ortho-trans content increased, contrarily to the trend observed for XLPE and XLPE_VE, for which no statistically significant changes in αo, αa and αb contents were detected. The XLPE_VE specimens underwent the least significant changes in the spectroscopic markers of micromorphology upon mechanical stress, probably due to their lower starting amorphous content. & 2015 Elsevier Ltd. All rights reserved.
1.
Introduction
failure (Brach del Prever et al., 2009; Bracco and Oral, 2011; James et al., 2009). Actually, during irradiation of UHMWPE,
It is well known that the oxidative degradation of ultra-high molecular weight polyethylene (UHMWPE) decreases its mechanical properties and represents the main cause of its n
Corresponding author. Tel.: þ39 051 6366864; fax: þ39 051 6366863. E-mail address: [email protected] (S. Affatato).
http://dx.doi.org/10.1016/j.jmbbm.2015.08.001 1751-6161/& 2015 Elsevier Ltd. All rights reserved.
free radicals are produced through the radiolytic cleavage of C–H and C–C bonds. They react with oxygen and trigger a complex cascade of reactions that lead to the formation of
journal of the mechanical behavior of biomedical materials 53 (2016) 40 –48
peroxy free radicals, hydroperoxides, and ultimately, carbonyl species, (mainly ketones, esters, and carboxylic acids). Thus, oxidation is accompanied by chain scission and reduction of the molecular weight of the polymer, possible recrystallisation, increase in stiffness and embrittlement, deteriorating the UHMWPE mechanical properties. The consequent increase in the wear debris production leads to biological responses that may cause osteolysis and implant loosening. The continuous and evolutive change from conventional UHMWPE to highly cross-linked polyethylenes has allowed many possible material combinations in the design of sliding couples for artificial hip joints. Highly cross-linked polyethylenes (XLPEs) were developed using high-dose irradiation (50– 100 kGy) to decrease the wear rate of UHMWPE (McKellop et al., 1999; Kurtz et al., 1999). Irradiation causes crosslinking in the amorphous phase of the UHMWPE, but also initiates the formation of free radicals in the crystalline phase; unable to recombine due to structural limitations, they become trapped for long periods of time (Oral et al., 2011; Singh 1999; Jahan et al., 2001; Micheli et al., 2012). These residual free radicals are believed to migrate to the crystalline/amorphous interface and cause oxidative degradation in the material through the above mentioned cascade of reactions with oxygen (Li et al., 2012; Bracco et al., 2007; Sobieraj and Rimnac, 2009; Mossman, 2008; Corcoran et al., 2009). The complete elimination of free radicals is difficult due to the limited mobility of these radicals within the crystalline structure of the polymer. Post-irradiation re-melting has been proposed to allow the recombination of free radicals trapped within the crystalline regions, with a consequent improvement in the oxidation and wear resistance (McKellop et al., 1999); unfortunately, remelting results in a decrease of PE crystallinity and thus in a loss of strength (Oral et al., 2006). Another strategy to retard oxidation consists in the addition of suitable stabilising additives, whose role is to decrease the reactivity of the radical species, to slow the oxidation processes and to preserve the PE chemical, physical and mechanical properties, i.e. to prolong its lifetime (Zweifel, 1998; Kurtz et al., 1999). Vitamin E seems to be the ideal candidate, being already employed as a natural antioxidant in the physiological processes of the human body (Bracco et al., 2007). The rationale for using this additive was the improvement of the oxidation resistance of irradiated UHMWPE as well as its fatigue strength, using an alternative to post-irradiation remelting (Bracco and Oral, 2011; Oral et al., 2008, 2010). Two methods may be used to incorporate vitamin E into PE, i.e. before consolidation and crosslinking or after irradiation. Besides prevention of oxidative degradation, the incorporation of vitamin E into the microstructure of PE components has been reported to determine a reduced incidence of fatigue crack, a reduced biological response and osteolytic potential induced by wear particle (Azzi and Stocker, 2000; Oral and Muratoglu, 2011). It is unclear if antioxidants improve UHMWPE durability in vivo by reducing oxidation and show less abrasion in the long-term. To date, there are no long-term clinical data available comparing XLPE with vitamin E stabilized UHMWPE/XLPE in hip arthroplasty liners; an in vivo evaluation of PE oxidation is not possible
41
(Kurtz et al., 2009; Lerf et al., 2010). At this moment, a longterm in vivo randomised controlled trial has been only designed to determine wear, bone mineral density, functional outcome and survival (Van der Veen et al., 2012). Salemyr et al. (2015) have carried out a prospective, randomized, controlled trial aimed at comparing a new vitamin-E diffused XLPE liner and a conventional XLPE liner, at a two-years followup; they have reported a lower initial head penetration and lower superior and medial wear for the former. However, they have acknowledged that a longer followup is necessary to demonstrate lower overall head penetration. In laboratory-based studies, various attempts have been made to compare the wear behaviour of conventional PE and vitamin-E stabilised liners (Affatato et al., 2010, 2012). However, the application of these models to the clinical situation did not appear satisfactory. Discrepancies have been observed on the role of the vitamin E between the results of laboratory wear tests and clinical observations. Moreover, contradictory results have been reported for in vitro studies using joint simulators. Affatato et al. (2012) in in vitro wear tests found that the vitamin-E blended XLPE wore more than XLPE and conventional UHMWPE. Previous studies have shown a reduction in wear rate with the addition of vitamin E to highly cross-linked UHMWPE compared to conventional UHMWPE, but in many of them, it is not clear whether the reduction in wear rate was due to the addition of vitamin E or the high levels of cross-linking (Micheli et al., 2012; Oral et al., 2006; Wannomae et al., 2010). Bladen et al. (2013) have reported that the addition of vitamin E to nonirradiated UHMWPE did not affect the wear factor suggesting that in previous studies, the reduction in wear was more likely to be associated with the high levels of crosslinking. On the other hand, Teramura et al. (2009) have reported that the wear volume of vitamin E-blended UHMWPE tested with a knee joint simulator was 30% lower than that of virgin UHMWPE after 5 million cycles (Mc). In this context, the main goal of this study was to comparatively investigate the wear performance of vitamin E-stabilised XLPE, XLPE and conventional UHMWPE acetabular cups in a hip joint simulator study. To gain more insights into the role of vitamin E in the wear mechanism of acetabular cups, the samples were analysed by micro-Raman spectroscopy. This technique proved suitable to investigate on a molecular scale the morphology changes (i.e. in crystallinity, phases distribution, chain orientation, residual strain) induced by wear. Micro-Raman spectroscopy was preferred over other techniques of molecular characterisation due to its nondestructive character, in view of continuing the tests under more severe conditions (i.e. after accelerated ageing and in the presence of third-body wear particles).
2.
Materials and methods
The wear behaviour of three different batches of polyethylene acetabular cups (32-mm inner 50-mm outer dimensions; 5 specimens for each batch) coupled with 32-mm cobalt–chromium–molybdenum (CoCrMo) femoral heads was investigated using a hip joint simulator. In particular, four components of each batch run onto the simulator and,
42
journal of the mechanical behavior of biomedical materials 53 (2016) 40 –48
following a standardized procedure (Affatato et al., 2005), another three acetabular cups (one for each type of material used) were stored (non-loaded) in bovine calf serum to compensate for weight changes due to fluid absorption. All polyethylenes tested in this study were machined from polymer bars Chirulen GUR 1020 (Polymax, Adler, Milan, Italy). Cross-linked acetabular cups (XLPE) were firstly γ-ray irradiated to 75 kGy (710%), then thermally treated at 150 1C, in order to remove free radicals formed during irradiation. After these treatments, the cups were machined to their final shape. Similarly, Vitamin E-containing (0.1%), XLPE acetabular cups (hereinafter called XLPE_VE) were machined from a Vitamin E-blended UHMWPE bars (Polymax, Adler, Milan, Italy), after electron beam irradiation to 75 kGy (710%), followed by a thermal treatment at 150 1C under nitrogen for 12 h. All polyethylene acetabular cups were pre-soaked for four weeks prior the wear tests.
2.1.
Wear test details
Wear tests were performed using a new 12-station hip joint simulator (IORSynthe, Bologna, Italy). The improved software of this hip simulator allows applying the kinematics as per ISO 14242-1:2012. The test was carried out after validation of the hip joint simulator; the ability of the hip simulator to follow the kinematic inputs and outputs as recommended by ISO 14242-1:2012 was verified. With the versatility of the improved software, all degree of freedom (DOF) was monitored, aligned, and modified in order to obtain all kinematic inputs fixed by the international recommendations and also to validate wear rates between individual stations. After validation, the UHMWPE, XLPE and XLPE_VE acetabular cups (three specimens for each set) were tested for two Mc. The simulator utilised hydraulic actuators to apply the cyclic vertical compressive loads (oscillating between 300 and 3000 N applied perpendicular to the acetabular components as recommend by the International ISO 14242). The lubricant used was 25 vol% new born calf serum balanced with distilled water, with 0.2% sodium azide in order to retard bacterial growth and 20 mM EDTA (ethylenediaminetetracetic acid) to minimise precipitation of calcium phosphate. The mass loss of the cups was determined every 0.4 million cycles using a new microbalance (SARTORIUS CUBIS MSE 225S-000-DU, Goettingën, Germany) with a sensitivity of 0.01 mg and an uncertainty of 0.01 mg. Each mass measurement was repeated three times and the average mass was used for calculations. Wear trend was determined from the mass loss of each acetabular cup, corrected by acetabular soak control; the wear rates, calculated from the steady-state slopes of the mass loss versus number of cycles lines, were obtained using least squares linear regression. The mass loss data were analysed using a nonparametric Kruskall–Wallis (K–W) test and a least significance difference as post-hoc test. Statistical significance was set at po0.05.
2.2.
Micro-Raman analyses
The PE components that showed the highest mass losses (two UHMWPE, one XLPE and one XLPE_VE) were analysed by micro-Raman spectroscopy since they were expected to show
the most significant changes on a molecular scale. The unworn control samples were characterised as well. MicroRaman spectra were obtained using a Jasco NRS-2000C instrument with a microscope of 50 magnification. All the spectra were recorded in back-scattering conditions with 5 cm 1 spectral resolution using the 532 nm Green Diode Pumped Solid State (DPSS) Laser Driver (RgBLase LLC, USA) with a power of ca. 20 mW. A 160 K cooled digital CCD (Spec-10: 100B, Roper Scientific Inc.) was used as a detector. A confocal pinhole with an aperture diameter of 200 μm was placed in the optical circuit to obtain signals from a limited in-depth region. All the Raman measurements were performed in a fully non-destructive way, without any sample manipulation. The worn components were analysed near the centre of the articulating surface, which appeared the most damaged area according to previously reported data (Bertoluzza et al., 2000; Affatato et al., 2003); twelve spectra at least were recorded. The same number of spectra was recorded on the unworn components. At a first approximation, PE can be considered as composed by three different phases (Strobl and Hagedorn, 1978): a crystalline phase (with the crystals being primarily orthorhombic in structure (Taddei et al., 2008, 2011)), a melt-like amorphous phase and an anisotropic disordered phase (i.e. a “third phase” with a prevailing trans conformation of the chains). Analogously to a previous study (Tozzi et al., 2014), we evaluated the effects of wear testing on the PE structure through several spectroscopic markers proposed in the literature (Strobl and Hagedorn, 1978; Lagaron et al., 1999). The fraction of orthorhombic (αo), amorphous (αa) and intermediate anisotropic disordered (αb) phases was calculated from the relative intensities of selected Raman bands, according to Strobl and Hagedorn (1978): αo ¼
A1414 0:46 A1295þ1305
ð1Þ
αa ¼
A1080 0:79 A1295þ1305
ð2Þ
αb ¼ 1 ðαo þ αa Þ
ð3Þ
where A1414 and A1080 are the areas of the Raman bands at 1414 and 1080 cm 1, respectively; A1295þ1305 is the area of the internal standard band group. The A1080 band area was determined trough curve fitting analysis by means of a commercial software (Opus 5.0 from Bruker Optik GmbH, Germany). Prior to curve fitting, the original spectra were baseline corrected. The Raman components were described as linear combinations of Gaussian and Lorentzian functions and curve fitting was performed using the LevenbergMarquardt algorithm. According to previous studies (Taddei et al., 2011; Kurelec et al., 2000), the occurrence of an orthorhombic to monoclinic phase transformation was evaluated through the I1414/ (I1440þI1460) intensity ratio, where I1414, I1440 and I1460 were calculated as peak heights. According to Lagaron et al. (1999), the fraction of all-trans sequences was evaluated through the following ratio: all trans ¼
A1130 0:80 A1295þ1305
ð4Þ
journal of the mechanical behavior of biomedical materials 53 (2016) 40 –48
where A1130 is the area of the Raman band at about 1130 cm 1, which has been associated with all-trans C–C bonds located both in the crystalline and amorphous phases. The fraction of all-trans sequences in an orthorhombic environment (i.e. ortho-trans) was calculated as ortho trans ¼
1:78 A1414 A1130
ð5Þ
The occurrence of orientation upon wear tests was assessed trough the A1130/A1060 area ratio. The 1060 cm 1 band area was calculated by curve fitting, according to the same procedure above reported. The residual strain of polyethylene was evaluated through the full-width at halfmaximum (FWHM) of the bands at about 1130 and 1060 cm 1.
3.
Results
All the prosthetic hip specimens completed the planned two Mc. As shown in Fig. 1, the XLPE_VE acetabular cups wore more than the XLPE ones, although the results of the posthoc test (Table 1) did not disclose any significant difference between the two sets of cups at any number of cycles. The XLPE_VE acetabular cups maintained a lower mass loss (although not statistically significant, Table 1) than the UHMWPE ones during the whole test. A statistically significant difference (p¼ 0.034) was observed between the XLPE and UHMWPE acetabular cups run onto the simulator for 1.2 and 2 million cycles (Table 1). No statistically significant differences were observed between the three sets of polyethylene cups using the K–S statistical test at the other number of cycles. Fig. 2 reports the micro-Raman spectra (normalised to the A1295þ1305 area) of the control unworn UHMWPE, XLPE and XLPE_VE acetabular cups, in the CH2 bending, CH2 twisting and C–C stretching regions; the assignments to the different phases have been given according to the literature (Mutter et al., 1993). As can be easily seen, they are not perfectly coincident, especially with regards to the bands at about 1440 and 1460 cm 1 (CH2 bending region) and 1160 cm 1 (C–C stretching region). The band at 1414 cm 1, assignable to the orthorhombic crystalline structure, showed nearly the same intensity in the
43
three spectra; accordingly, the αo orthorhombic content in the three sets of acetabular cups was nearly the same and no significantly differences were detected (p ¼0.056, Fig. 4A). The band at 1080 cm 1, assignable to the amorphous phase, appeared slightly weaker in the spectrum of the XLPE_VE control cup (Fig. 2); actually, the statistical analysis showed for this specimen a significantly lower αa amorphous content value than for UHMWPE and XLPE (UHMWPE versus XLPE_VE: po0.001; XLPE versus XLPE_VE: p ¼0.001, Fig. 3A). No significant differences were detected between the αb third phase content values of the three sets of cups (p ¼0.54, Fig. 3A). Fig. 3B shows statistically significant differences between UHMWPE and XLPE (in all-trans and ortho-trans contents, A1130/A1060; po0.001) as well as between UHMWPE and XLPE_VE (in all-trans and ortho-trans contents, A1130/ A1060 (po0.001) and I1414/(I1440þI1460) p ¼0.045). From the data reported in Fig. 3A and B, it can be easily seen that the spectroscopic markers corresponding to XLPE and XLPE_VE control acetabular cups had an associated standard deviation higher than for UHMWPE. For most spectroscopic markers, the standard deviation associated to XLPE_VE was the highest. The differences in the all-trans and ortho-trans contents as well as in the A1130/A1060 ratio (Fig. 3B) reflect the differences in intensity of the band at about 1130 cm 1, which is significantly higher in UHMWPE than in XLPE and XLPE_VE (Fig. 2). The I1414/(I1440þI1460) ratio is significantly higher in XLPE_VE (Fig. 3B), as expectable on the basis of the significantly lower intensity of the bands at 1440 and 1460 cm 1, due to the higher amorphous content present in this sample, also revealed by the higher αa value (Fig. 3A). The three sets of unworn samples were characterised by significant differences in the FWHM values attained by both the 1130 and 1060 cm 1 bands (Fig. 3C), suggesting that crosslinking and addition of vitamin E modified the residual strain in the material. The data reported in Figs. 4 and 5 showed the effects of wear on the micromorphology of the three sets of cups, as investigated through the αo, αa, αb, all-trans and ortho-trans contents, I1414/(I1440þI1460) and A1130/A1060 ratios as well as the
Fig. 1 – Wear behaviour and regression coefficient for the different polyethylenes tested.
44
journal of the mechanical behavior of biomedical materials 53 (2016) 40 –48
Table 1 – Cumulative mass loss (mean7standard deviation) for the three sets of PE acetabular cups tested. Values and statistical analysis performed using a Kruskall–Wallis nonparametric test. Mean7standard deviation
K–W test (p-value)
Cycles [Mc]
UHMWPE
XLPE
XLPE_VE
0.4 0.8 1.2 1.6 2
5.270.5 13.170.6 24.571.7 27.971.9 35.473.3
1.870.7 2.771.2 3.371.5 4.772.7 6.774.0
1.871.0 3.871.5 8.975.5 11.577.5 16.178.2
0.066 0.066 0.039 0.393 0.039
Post-hoc test (p-value) UHMWPE versus XLPE
UHMWPE versus XLPE_VE
XLPE versus XLPE_VE
0.54 0.33 0.034 0.33 0.034
0.54 0.76 0.76 0.76 0.76
1.00 0.76 0.76 0.76 0.76
NS ¼not significant (p value40.05); K–W ¼Kruskall–Wallis; Mc ¼millions cycles.
Fig. 2 – Micro-Raman spectra (normalised to the A1295þ1305 area) of the control unworn UHMWPE (black), XLPE (blue) and XLPE_VE (red) acetabular cups, in the CH2 bending, CH2 twisting and C–C stretching regions; the assignments to the amorphous (A) and crystalline (C) phases have been given according to the literature. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
FWHM1130 and FWHM1060. Upon wear, the UHMWPE acetabular cups showed a significant increase of αo (p¼ 0.041 for #2) αa (p ¼0.010 and o0.001 for #1 and #2, respectively, Fig. 4) and ortho-trans contents (po0.001 and p ¼0.001 for #1 and #2, respectively, Fig. 5) and a significant decrease of αb (p¼ 0.001 for #2, Fig. 4) and all-trans contents (po0.001 and p ¼0.04 for #1 and #2, respectively, Fig. 5), I1414/(I1440þI1460) (p ¼0.008 for #1, Fig. 5) and A1130/A1060 ratios (p¼ 0.045 and 0.012 for #1 and #2, respectively, Fig. 5). More significant changes in the FWHM1130 and FWHM1060 were observed for the sample #2, suggesting that wear has modified the residual strain in this component (Fig. 5). If compared with UHMWPE, the XLPE and XLPE_VE acetabular cups showed a different trend of the spectroscopic
markers upon wear and appeared similar each other: the alltrans content significantly increased (p¼ 0.004 and po0.001 for XLPE and XLPE_VE, respectively, Fig. 5) as well as the A1130/A1060 ratio (po0.001 for both XLPE and XLPE_VE, Fig. 5), while the ortho-trans content decreased (p ¼0.0017 and p ¼0.004 for XLPE and XLPE_VE, respectively, Fig. 5). No significant changes were observed for αo, αa and αb contents.
4.
Discussion
Due to contradicting results on the in vitro wear performance of vitamin E-containing PE (Affatato et al., 2010, 2012; Wannomae et al., 2010; Bladen et al., 2013; Teramura et al., 2009), a hip joint
journal of the mechanical behavior of biomedical materials 53 (2016) 40 –48
45
Fig. 3 – Values (average7standard deviation) of αo, αa and αb contents (A), all-trans and ortho-trans contents, I1414/ (I1440þI1460) and A1130/A1060 ratios (B), FWHM1130 and FWHM1060 (C), as obtained from the spectra of the control unworn UHMWPE, XLPE and XLPE_VE acetabular cups. Arrows with asterisks indicate statistically significant differences (po0.05).
Fig. 4 – Values (average7standard deviation) of αo, αa and αb contents, as calculated from the spectra of the control unworn and worn UHMWPE, XLPE and XLPE_VE acetabular cups. Arrows with asterisks indicate statistically significant differences with respect to control cups (po0.05).
simulator study has been carried out. Using a 12-station hip simulator, the wear performance of three different sets of PE acetabular cups coupled with CoCrMo femoral heads was evaluated for two million cycles; we asked whether the addition of vitamin E to XLPE would improve the wear behaviour of such material with respect to the conventional UHMWPE and XLPE. The results of this investigation clearly showed a reduced wear for the XLPE and the XLPE_VE acetabular cups with respect to the conventional UHMWPE components. However, a statistically significant difference was observed only between the XLPE and UHMWPE acetabular cups run onto the simulator for 1.2 and 2 million cycles (Table 1). No statistically significant differences were observed between the three sets of polyethylene cups using the K–S statistical test at the other number of cycles.
To gain insights into these results on a molecular scale, micro-Raman spectroscopic measurements on control and simulator tested UHMWPE, XLPE and XLPE_VE acetabular cups were carried out. With regards to the control cups, the XLPE_VE unworn cup showed a statistically significant lower αa amorphous content (Fig. 3A) than the other unworn cups. With regards to the all-trans and orthotrans contents as well as the A1130/A1060 ratio, XLPE and XLPE_VE control cups did not appear significantly different each other, while showed statistically significant differences with UHMWPE (Fig. 3B). On the basis of the obtained results, it may be affirmed that crosslinking and post-irradiation melting did not sensibly alter the distribution of the crystalline and amorphous phases (i.e. orthorhombic, amorphous and third intermediate phase),
46
journal of the mechanical behavior of biomedical materials 53 (2016) 40 –48
Fig. 5 – Values (average7standard deviation) of all-trans and ortho-trans contents, I1414/(I1440þI1460) and A1130/A1060 ratios, FWHM1130 and FWHM1060, as calculated from the spectra of the control unworn and worn UHMWPE, XLPE and XLPE_VE acetabular cups. Arrows with asterisks indicate statistically significant differences with respect to control cups (po0.05). since their content in XLPE control cup was nearly the same as in UHMWPE (Fig. 3A). However, these processes induced structural rearrangements that affected the all-trans and ortho-trans contents (the former decreased, the latter increased, Fig. 3B), as well as the chain orientation (as revealed by the decrease of the A1130/A1060 ratio, Fig. 3B). Actually, the bands at 1130 and 1060 cm 1 have different vibrational symmetries; if the molecules are oriented in a preferred direction, the 1130 cm 1 band has been reported to become stronger with respect to the 1060 cm 1 band (Pigeon et al., 1991). The presence of vitamin E during cross-linking and postirradiation did not sensibly affect the value of the latter spectroscopic markers, while determined a decrease of the amorphous phase content. The differences observed in the control acetabular cups may have determined the differences in the wear behaviour
of the three sets of acetabular cups. Actually, it is well recognised that the mechanical properties of PE depend on the distribution of crystalline and non-crystalline phases as well as molecular orientation, besides molecular weight and degree of cross-linking. The latter aspect may explain the trend of the mass loss which was found to decrease along the series UHMWPE4XLPE_VE4XLPE, although significant differences were observed only between UHMWPE and XLPE at 1.2 and 2 Mc. It is well recognised that under multidirectional stress conditions an anisotropic PE surface leads to a decreased resistance to adhesive wear due to strain-softening and weakening. These adverse effects are mitigated by cross-linking, due to the increased number of intralamellar and inter-lamellar covalent bonds. Therefore, the results of the present test confirmed the positive effects of cross-linking on wear resistance.
journal of the mechanical behavior of biomedical materials 53 (2016) 40 –48
The morphological differences observed by micro-Raman spectroscopy in the control acetabular cups determined different wear behaviours at molecular level (i.e. different trends of the spectroscopic markers upon wear). As stressed above, the main spectroscopic differences were observed by comparing control UHMWPE versus XLPE and XLPE_VE (Fig. 3); actually, XLPE and XLPE_VE appeared quite similar each other. As expectable, they underwent similar changes upon wear, which appeared significantly different (i.e. opposite in several aspects) to those experienced by UHMWPE (Figs. 4 and 5). Wear testing proved to induce a stress-induced crystallisation in UHMWPE, in agreement with previous studies (Affatato et al., 2005, 2013); actually, the αo content appeared to significantly increase upon wear (Fig. 4), as a result of the near-surface orientation of the crystalline lamellae observed in the wear-tested acetabular liners (Edidin et al., 1999). The αo increase upon wear in XLPE and XLPE_VE did not appear significant, in agreement with previous studies (Affatato et al., 2005); actually, the thickness of the plasticity-induced damage layer that precedes abrasive wear has been reported to be lower in XLPE. Also the αa amorphous content appeared to significantly increase upon wear in UHMWPE (Fig. 4); both αa and αo increased at the expense of αb, which decreased upon wear. At the same time, the all-trans content decreased while the ortho-trans content increased (Fig. 5). XLPE and XLPE_VE showed an opposite trend upon wear: actually, the ortho-trans content decreased, while the all-trans content increased (Fig. 5); no statistically significant changes in αo, αa and αb were detected. With regards to XLPE_VE, it must also be observed that vitamin E has been suggested to be present within the amorphous phase of PE (Mallégol et al., 2001) and for its high affinity for the polymer, it inhibits the strain-induced crystallisation phenomenon (Okubo et al., 2009). Also molecular orientation effects appeared different in UHMWPE versus XLPE and XLPE_VE. In the former, the A1130/A1060 ratio decreased upon wear (Fig. 5), while increased in the crosslinked specimens (i.e. XLPE and XLPE_VE). The spectroscopic results disclosed a different wear mechanism between UHMWPE and cross-linked PEs. It must be stressed that upon wear, the XLPE_VE acetabular cups appeared the least affected by mechanical stress; actually, this kind of specimens underwent the least significant changes of the various spectroscopic markers investigated (Figs. 4 and 5). Probably, these slighter morphological alterations induced by wear may be related to its lower starting αa content (Fig. 3A). Actually, a lower content of the amorphous phase has been related to a higher resistance to creep deformation, but to a lower capacity of shape recovery after removal of the applied load (Takahashi et al., 2014).
5.
Conclusions
The acetabular cups studied in this work showed a mass loss upon wear testing decreasing along the series UHMWPE4XLPE_VE4XLPE, although statistically significant differences were found only between the mass losses of XLPE and UHMWPE at 1.2 and 2 million cycles. The mass loss data were explained in relation to molecular morphology, as
47
investigated by non-destructive micro-Raman spectroscopy. This technique allowed to disclose a different wear behaviour of the UHMWPE and cross-linked PE specimens (i.e. an opposite trend of several spectroscopic markers); the vitamin E-blended specimens underwent the least significant micromorphological changes upon mechanical stress.
Acknowledgements The authors would like to thank Sami Abdel Jaber for his help during the set-up and Tommaso Vannini for his help with statistical analyses. Thanks are due also to Riccardo Verga (Research & Development, Adler Ortho S.r.l, Milan, Italy) for his supplying of the samples. The authors would like to express their gratitude to the Synthesis (Florence, Italy) for using the simulator. This work was partially supported by the Italian Program of Donation for Research “5 per mille”, (Cardinis 5136) year 2011.
r e f e r e nc e s
Affatato, S., Modena, E., Carmignato, S., Taddei, P., 2013. The use of Raman spectroscopy in the analysis of UHMWPE unicondylar bearing systems after run on a force and displacement control knee simulators. Wear 297, 781–790. Affatato, S., Bersaglia, G., Emiliani, D., Foltran, I., Taddei, P., Reggiani, M., et al., 2003. The performance of gamma- and EtO-sterilised UHMWPE acetabular cups tested under severe simulator conditions. Part 2: wear particle characteristics with isolation protocols. Biomaterials 24, 4045–4055. Affatato, S., Bersaglia, G., Rocchi, M., Taddei, P., Fagnano, C., Toni, A., 2005. Wear behaviour of cross-linked polyethylene assessed in vitro under severe conditions. Biomaterials 26, 3259–3267. Affatato, S., Zavalloni, M., Spinelli, M., Costa, L., Bracco, P., Viceconti, M., 2010. Long-term in-vitro wear performance of an innovative thermo-compressed cross-linked polyethylene. Tribol. Int. 43, 22–28. Affatato, S., Bracco, P., Costa, L., Villa, T., Quaglini, V., Toni, A., 2012. In vitro wear performance of standard, crosslinked, and vitamin-E-blended UHMWPE. J. Biomed. Mater. Res. A 100, 554–560. Azzi, A., Stocker, A., 2000. Vitamin E: non-antioxidant roles. Prog. Lipid Res. 39, 231–255. Bertoluzza, A., Fagnano, C., Rossi, M., Tinti, A., Cacciari, G.L., 2000. Micro-Raman spectroscopy for the crystallinity characterization of UHMWPE hip cups run on joint simulators. J. Mol. Struct. 521, 89–95. Bladen, C.L., Teramura, S., Russell, S.L., Fujiwara, K., Fisher, J., Ingham, E., et al., 2013. Analysis of wear, wear particles, and reduced inflammatory potential of vitamin E ultrahighmolecular-weight polyethylene for use in total joint replacement. J. Biomed. Mater. Res. B 101 B, 458–466. Bracco, P., Oral, E., 2011. Vitamin E-stabilized UHMWPE for total joint implants: a review. Clin. Orthop. Relat. Res. 469, 2286–2293. Bracco, P., Brunella, V., Zanetti, M., Luda, M.P., Costa, L., 2007. Stabilisation of ultra-high molecular weight polyethylene with vitamin E. Polym. Degrad. Stab. 92, 2155–2162. Brach del Prever, E.M., Bistolfi, A., Bracco, P., Costa, L., 2009. UHMWPE for arthroplasty: past or future?. J. Orthop. Traumatol. 10, 1–8.
48
journal of the mechanical behavior of biomedical materials 53 (2016) 40 –48
Corcoran, P.L., Biesinger, M.C., Grifi, M., 2009. Plastics and beaches: a degrading relationship. Mar. Pollut. Bull. 58, 80–84. Edidin, A.A., Pruitt, L., Jewett, C.W., Crane, D.J., Roberts, D., Kurtz, S.M., 1999. Plasticity-induced damage layer is a precursor to wear in radiation-cross-linked UHMWPE acetabular components for total hip replacement. J. Arthroplast. 14, 616–627. Jahan, M.S., King, M.C., Haggard, W.O., Sevo, K.L., Parr, J.E., 2001. A study of long-lived free radicals in gamma-irradiated medical grade polyethylene. Radiat. Phys. Chem. 62, 141–144. James S.P., Oldinski R. , Zhang M., Schwartz H., 2009. UHMWPE Biomaterials Handbook. Kurelec, L., Rastogi, S., Meier, R.J., Lemstra, P.J., 2000. Chain mobility in polymer systems: on the borderline between solid and melt. 3. Phase transformations in nascent ultra high molecular weight polyethylene reactor powder at elevated pressure as revealed by in-situ Raman spectroscopy. Macromolecules 33, 5593–5601. Kurtz, S.M., Muratoglu, O.K., Evans, M., Edidin, A.A., 1999. Advances in the processing, sterilization, and crosslinking of ultra-high molecular weight polyethylene for total joint arthroplasty. Biomaterials 20, 1659–1688. Kurtz, S.M., Dumbleton, J., Siskey, R.S., Wang, A., Manley, M., 2009. Trace concentrations of vitamin E protect radiation crosslinked UHMWPE from oxidative degradation. J. Biomed. Mater. Res. A 90, 549–563. Lagaron, J.M., Dixon, N.M., Reed, W., Pastor, J.M., Kip, B.J., 1999. Morphological characterisation of the crystalline structure of cold-drawn HDPE used as a model material for the environmental stress cracking (ESC) phenomenon. Polymer 40, 2569–2586. Lerf, R., Zurbru¨gg, D., Delfosse, D., 2010. Use of vitamin E to protect cross-linked UHMWPE from oxidation. Biomaterials 31, 3643–3648. Li, C.S., Zhan, M.S., Huang, X.C., Zhou, H., 2012. Degradation behavior of ultra-high molecular weight polyethylene fibers under artificial accelerated weathering. Polym. Test. 31, 938–943. Malle´gol, J., Carlsson, D.J., Descheˆnes, L., 2001. Post-γ-irradiation reactions in vitamin E stabilised and unstabilised HDPE. Nucl. Instrum. Methods Phys. Res. B 185, 283–293. McKellop, H., Shen, F.W., Lu, B., Campbell, P., Salovey, R., 1999. Development of an extremely wear-resistant ultra high molecular weight polyethylene for total hip replacements. J. Orthop. Res. 17, 157–167. Micheli, B.R., Wannomae, K.K., Lozynsky, A.J., Christensen, S.D., Muratoglu, O.K., 2012. Knee simulator wear of vitamin E stabilized irradiated ultrahigh molecular weight polyethylene. J. Arthroplast. 27, 95–104. Mossman, S., 2008. Conservation of plastics: materials science, degradation and preservation. Nature 455, 288–289. Mutter, R., Stille, W., Strobl, G., 1993. Transition regions and surface melting in partially crystalline polyethylene: a Raman spectroscopic study. J. Polym. Sci. B: Polym. Phys. 31, 99–105. Okubo, Y., Teramura, S., Niwa, Y., Ibaraki, K., Murata, K., Hyon, S. H., et al., 2009. Strain-induced crystallization and orientation of vitamin E-blended ultrahigh molecular weight polyethylene. Biomed. Mater. Eng. 19, 431–439. Oral, E., Muratoglu, O.K., 2011. Vitamin E diffused, highly crosslinked UHMWPE: a review. Int. Orthop. 35, 215–223. Oral, E., Christensen, S.D., Malhi, A.S., Wannomae, K.K., Muratoglu, O.K., 2006. Wear resistance and mechanical properties of highly cross-linked, ultrahigh-molecular weight polyethylene doped with vitamin E. J. Arthroplast. 21, 580–591. Oral, E., Malhi, A.S., Muratoglu, O.K., 2006. Mechanisms of decrease in fatigue crack propagation resistance in irradiated and melted UHMWPE. Biomaterials 27, 917–925.
Oral, E., Malhi, A.S., Wannomae, K.K., Muratoglu, O.K., 2008. Highly cross-linked ultrahigh molecular weight polyethylene with improved fatigue resistance for total joint arthroplasty. Recipient of the 2006 Hap Paul Award. J. Arthroplast. 23, 1037–1044. Oral, E., Ghali, B.W., Rowell, S.L., Micheli, B.R., Lozynsky, A.J., Muratoglu, O.K., 2010. A surface crosslinked UHMWPE stabilized by vitamin E with low wear and high fatigue strength. Biomaterials 31, 7051–7060. Oral, E., Ghali, B.W., Muratoglu, O.K., 2011. The elimination of free radicals in irradiated UHMWPEs with and without vitamin e stabilization by annealing under pressure. J. Biomed. Mater. Res. B 97 B, 167–174. Pigeon, M., Prud’homme, R.E., Pezolet, M., 1991. Characterization of molecular-orientaiton in polyethyleen by Ramanspectroscopy. Macromolecules 24, 5687–5694. Salemyr, M., Muren, O., Ahl, T., Bode´n, H., Chammout, G., Stark, A., et al., 2015. Vitamin-E diffused highly cross-linked polyethylene liner compared to standard liners in total hip arthroplasty. A randomized, controlled trial. Int. Orthop. 39 (8), 1499–1505. Singh, A., 1999. Irradiation of polyethylene: some aspects of crosslinking and oxidative degradation. Radiat. Phys. Chem. 56, 375–380. Sobieraj, M.C., Rimnac, C.M., 2009. Ultra high molecular weight polyethylene: mechanics, morphology, and clinical behavior. J. Mech. Behav. Biomed. Mater. 2, 433–443. Strobl, G.R., Hagedorn, W., 1978. Raman spectroscopic method for determining the crystallinity of polyethylene. J. Polym. Sci.: Polym. Phys. Ed. 16, 1181–1193. Taddei, P., Affatato, S., Rocchi, M., Fagnano, C., Viceconti, M., 2008. The effects of irradiation and EtO-treatment on ultrahigh molecular weight polyethylene acetabular cups following accelerated aging: degradation of mechanical properties and morphology changes during hip simulator tests. J. Mol. Struct. 875, 254–263. Taddei, P., Di Foggia, M., Affatato, S., 2011. Raman characterization of conventional and cross-linked polyethylene in acetabular cups run on a hip joint simulator. J. Raman Spectrosc. 42, 1344–1352. Takahashi Y., Masaoka T., Pezzotti G., Shishido T., Tateiwa T., Kubo K., et al., 2014. Highly cross-linked polyethylene in total hip and knee replacement : spatial distribution of molecular orientation and shape recovery behavior, BioMed. Research International, Vol. 2014 (2014), 1-13, http://dx.doi.org/10.1155/ 2014/808369. Teramura, S., Sakoda, H., Terao, T., Fujiwara, K., Kawai, K., Tomita, N., 2009. Reduction of wear volume from accelerated aged UHMWPE knee components by the addition of vitamin E. J. Biomech. Sci. Eng. 4, 589–596. Tozzi, S., Modena, E., Falcioni, S., Sudanese, A., Affatato, S., Taddei, P., 2014. The effects of contact area and applied load on the morphology of in vitro worn ultra-high molecular weight knee prostheses: a micro-Raman and gravimetric study. J. Raman Spectrosc. 45, 781–787. Van der Veen, H.C., van den Akker-Scheek, I., Bulstra, S.K., van Raay, J.J., 2012. Wear, bone density, functional outcome and survival in vitamin E-incorporated polyethylene cups in reversed hybrid total hip arthroplasty: design of a randomized controlled trial. BMC Musculoskelet. Disord. 13, 178. Wannomae, K.K., Christensen, S.D., Micheli, B.R., Rowell, S.L., Schroeder, D.W., Muratoglu, O.K., 2010. Delamination and adhesive wear behavior of α-tocopherol-stabilized irradiated ultrahigh-molecular-weight polyethylene. J. Arthroplast. 25, 635–643. Zweifel, H., 1998. Stabilization of Polymeric Materials. Springer, Berlin, Germany.