Biomaterials 26 (2005) 633–643

ARTICLE IN PRESS

*Correspondin

334-567.

E-mail addres

0142-9612/$ - see

doi:10.1016/j.bio

Analysis and evaluation of a biomedical polycarbonate urethanetested in an in vitro study and an ovine arthroplasty model.

Part II: in vivo investigation

Imran Khana, Nigel Smithb, Eric Jonesc, Dudley S. Finchd, Ruth Elizabeth Camerona,*aDepartment of Materials Science and Metallurgy, University of Cambridge, New Museums Site, Pembroke Street, Cambridge CB2 3QZ, UK

bFormerly of Stryker Howmedica Osteonics, Fairfield Avenue, Staines, UKcStryker Europe (CASA group), Limerick, Ireland

dDepartment of Mechanical Engineering, University of Colorado, at Boulder, 427 UCB, Boulder, CO 80309-0427, USA

Received 26 September 2003; accepted 24 February 2004

Abstract

The polyurethane (PU) elastomer Corethane 80A (Corvita) is being considered as the acetabular bearing material in a novel total

replacement hip joint. Its biostability was investigated in vitro (Analysis and evaluation of a biomedical polycarbonate urethane

tested in an in vitro study and an ovine arthroplasty model. Part I: material selection and evaluation, Biomaterials, in press) together

with three other commercially available biomedical PUs: Pellethane 2363-80A (DOW Chemical), a polyhexamethylene oxide based

PU, PHMO-PU (CSIRO, not supplied as a commercial product) and ChronoFlex AL-80A (CardioTech). From the in vitro studies,

Corethane 80A displayed the best overall resistance to hydrolysis, ESC, MIO and calcification, followed by ChronoFlex 80A and

PHMO-PU, with Pellethane 80A being the least stable. Building on the in vitro investigation, the follow-up in vivo study (reported

here) assessed Corethane 80A as the bearing layer in a prototype compliant layer acetabular cup, in a fully functioning ovine total

hip arthoplasty (THA) model. PU degradation in the retrieved cups was analysed using a range of analytical and physical-testing

methods including mechanical testing, differential scanning calorimetry, Fourier transform infrared spectroscopy and

environmental scanning electron microscopy. The Corethane 80A functioned well in the THA model, with the bearing surfaces

of the retrieved hip cups showing no significant evidence of biodegradation or wear damage after 3 years in vivo. The findings in this

study provide compelling evidence for the biostability and effectiveness of acetabular cups incorporating a Corethane 80A

compliant bearing layer.

r 2004 Elsevier Ltd. All rights reserved.

Keywords: Hip replacement prostheses; Polyurethane; Elstomer; Compliance; Biocompatibility; Animal model

1. Introduction

The long-term survival of total hip replacement [1]can be influenced by a wide range of factors includingthe nature of the implant used, surgical technique [2–5],the condition of the host bone, post-operative care, andthe age, health and activity level of the patient. Inconventional acetabular components designed for totalhip arthroplasty, one of the most frequent failuremechanisms has been wear particle mediated osteolysis

g author. Tel.: +44-1223-334-324; fax: +44-1223-

s: recll@cam.ac.uk (R.E. Cameron).

front matter r 2004 Elsevier Ltd. All rights reserved.

materials.2004.02.064

[6–9]. Acetabular component failure is a common reasonfor revision surgery [10–23]. A prosthesis incorporatinga compliant layer with greater resistance to wear andin vivo degradation should significantly reduce the levelof wear particle mediated osteolysis and therebyimprove the long-term stability.Polyurethane (PU) elastomers have a unique combi-

nation of toughness, durability and flexibility, biocom-patibility and biostability that makes them suitablematerials for use in a diverse range of implantablemedical devices. Our in vitro studies assessed theresistance of Corethane 80A to the main degradationmechanisms observed in PUs: hydrolysis, environmentalstress cracking (ESC), metal ion oxidation (MIO) andcalcification.

ARTICLE IN PRESS

Fig. 1. Compliant layer acetabular cup for the ovine hip arthroplasty

model.

I. Khan et al. / Biomaterials 26 (2005) 633–643634

Corethane 80A, a commercially available polycarbo-nate urethane (PCU), presently branded as Bionate(produced under licence by Polymer Technology Group,Berkeley, CA), exhibited excellent resistance to thesedegradation mechanisms, implying that the materialcould provide a more robust bearing layer in aprototype compliant layer acetabular cup [1].In this second part of the two papers on material, we

report a series of tests on samples of Corethane 80Aretrieved from an in vivo trial using a fully functioningtotal hip arthoplasty (THA) with a Corethane 80Acompliant layer implanted into sheep, and makecomparisons with a series of in vitro cups stored at37�C in dry and PBS environments.The sheep proximal femur, although a different shape,

is a reasonable morphological analogue to the humanfemur, making the ovine model increasingly popular fororthopaedic conditions [24,25]. Sheep tolerate theprocedure well and bear weight early in the post-operative period, making this model more appropriatethan canine models [26–28].This paper reports on the biostability of the PU

material and forms part of a research programmedesigned to evaluate a soft layer bearing from itstheoretical design through to a practical demonstrationof the technology in an in vivo model of hiparthroplasty. This research is being reported from designconception through material testing to a practicalapplication of a novel acetabular cup design implantedinto an ovine hip arthroplasty model for 4 years. Theoverall objective was to develop a non-polyethylenebearing system that performs in a way analogous tonatural articular cartilage by limiting the wear debrisproduction by almost frictionless articulation and thusextending the useful lifetime of artificial joints [29–31].

Fig. 2. Steps required to manufacture a compliant layer acetabular

cup.

2. Experimental materials and procedures

The PU cup used in the trial comprises two layers: thesofter inner bearing layer, which is made from Cor-ethane 80A, and an outer shell made from a hardergrade PU, Corethane 75D (Fig. 1). The prototypecemented THA joint (acetabular and femoral compo-nents) is a scaled-down design suitable for sheep basedon Howmedica’s Exeter Hip System.The cups were manufactured in a two-part injection

moulding process [32], with the Corethane 75D shellmoulded first followed by the soft compliant bearinglayer on its inner surface (Fig. 2).Thirty-seven mature Merino wethers (3 years old),

with a mean weight of 55 kg were taken from an existingflock with a common genetic background and experi-ence, and underwent total hip arthroplasty [33].Implants were retrieved from groups of eight animalsafter 6 months, 1- and 2-year implantation times, with

the final group of eight animals scheduled to besacrificed after 4 years (May–June 2002, to be reportedlater). One animal from the 4-year trial died prematurelyafter 3 years from complications unrelated to implanta-tion, providing an additional cup for 3-year evaluation.A single ultra-high-molecular-weight polyethylene(UHMWPE) acetabular cup, of equivalent design wasincluded in the 1-year group for comparison.The total hip arthroplasty was carried out by Field

et al. [30]. The animals were contained in a paddock forthe duration of the trial, and were exercised by walkingfor 5 km on a daily basis. At appropriate time intervals,animals were sacrificed according to Institutional Ethics

ARTICLE IN PRESS

Fig. 3. Sampling retrieved PU acetabular cups.

I. Khan et al. / Biomaterials 26 (2005) 633–643 635

Committee guidelines. All retrieved specimens werephotographed (Fig. 3(a)) for the record, and the excisedfemora and acetabulae were radiographed using aFaxitron X-ray cabinet (Hewlett-Packard) in both theanterior–posterior and lateral views to examine theprosthesis–cement–bone interface. The excised hip jointswere then scanned using DEXA to assess changes inbone density around the femoral and acetabularcomponents. The tibiae were scanned to assess anyasymmetry in weight bearing between the left and rightlegs.Synovial capsular specimens were taken and exam-

ined histologically to exclude infection, to identifyevidence of local tissue response to the implant, and todetect the presence and number of particles usingdifferential staining and polarized light microscopy[34]. The tissue of the synovial capsule and adjacentsoft tissues were also examined histologically to detectpossible debris. Other tissues and organs such as theliver, spleen and lymphatic glands were assessed forabnormal tissue responses. Following histological ana-lysis, samples were shipped and stored in a 40% ethanolsolution prior to topographical and physicochemicalanalysis.Professor Donald Howie’s Group (Department of

Orthopaedic Surgery and Trauma, Royal AdelaideHospital, Australia) was responsible for animal care,operative procedures and post-operative managementand histological analysis.

Samples from the study were examined to determinethe effect of implantation time upon the final conditionof the compliant layer (CL) cup. The effects of in vivodegradation on the physicochemical properties of thecup material and in-service wear damage were investi-gated.The bearing surface of each retrieved sample was first

photographed and then analysed using optical stereo-microscopy and environmental scanning electron micro-scopy (ESEM). ESEM was performed using anElectroscan E3, ESEM. The whole cup was mountedon a large aluminium sample stub, reference markers ofaluminium tape were placed on the rim of the cup andthe whole assembly was placed on the sample stage.Cores (2.5mm diameter� 1.5mm deep) were taken

from two regions of the Corethane 80A bearing layer:region 1 near the hood, the area of maximum loadingduring the normal walking cycle; and region 2 on theopposite side of the cup, the area that experiences theleast loading (Fig. 3(b)). Cores were obtained by drillingthrough to the shell/bearing layer interface using abench top drill equipped with a 5mm diameter flat-tipped drill-bit. A custom made steel corer was then usedto remove cores from the Corethane 80A bearing layer(Fig. 3(c)) to give cylindrical cores with a ‘flat’ bearingsurface (Fig. 3(d)).Changes in the physicochemical properties of the

retrieved cores and cores from non-implanted wet (37�Cin phosphate buffered saline (PBS)) and dry (37�C in

ARTICLE IN PRESSI. Khan et al. / Biomaterials 26 (2005) 633–643636

anhydrous atmosphere) control samples were assessedusing a variety of analysis and testing techniques. Load-penetration was used to evaluate bulk changes inmechanical properties. Nanoindentation measuredchanges in micromechanical properties, while Fouriertransform infrared spectroscopy (FTIR, Perkin-ElmerSpectrum GX FTIR, equipped with a Specac GoldenGate Single Reflection Diamond ATR System) assessedchemical changes. Differential scanning calorimetry(DSC) was used to examine shifts in endotherms thatmight indicate microdomain ordering. Changes inmolecular weight were evaluated using gel permeationchromatography (GPC) was performed by RAPRA,Shrewsbury, UK, using a Polymer Laboratories GPC220 instrument equipped with a refractive index detector(with differential pressure and light scattering) and a setof two PLgel 5 mm MIXED-D Columns (300� 7.5mm).HPLC grade tetrahydrofuran (THF) with an addedantioxidant was used as the eluent. The analyses wererun at 30�C and a flow rate of 1.0mlmin�1. Theinstrument was calibrated using polystyrene standards(580–400 000). Surface roughness measurements weremade using WYKO surface profilimetry. All samples,including the wet controls, were dried for 24 h at roomtemperature (RT) and pressure before being analysed toeliminate hydration effects [1].

3. Results and discussion

3.1. Macroscopic observations

No significant physical or chemical degradation wasobserved after 3 years in vivo (Fig. 4). Host variablessuch as weight, age, mobility and implant durationproduced little difference and there was remarkableconsistency in the condition of the cups retrieved.At retrieval, observations of fractures in the Cor-

ethane 75D shell were rare, occurring on only oneoccasion. Fracture of the hard shell however, did notcause the compliant layer (Corethane 80A) to fail norresult in loosening of the cup. The articular surfaces

Fig. 4. Representative photographs of cups retrieved at 1 year along with en

the compliant layer can be seen in the compliant layer (black arrow), and o

were in excellent condition suggesting that there wasgood joint function over the implant period [35].Impingement wear on the cup hood and shell was a

frequent finding that may require design changes.ESEM and surface profilimetry showed minimal normalwear, supporting the clinical findings that revealed nosignificant soft tissue damage, wear particle generationor inflammatory tissue response [35].There was increased ordering of the hard microdo-

main in the CL, particularly in the high load bearingregion (region 1) that caused a small increase in stiffness.Although stress did appear to affect the microstructureand mechanical properties of the CL, it did not appearto affect the rate or amount of chemical degradation,and no significant surface or bulk chemical degradationwas observed.There was no evidence of biodegradation such as

environmental stress cracking, which can result in thematerial becoming more opaque (Fig. 5). In general,burnishing was noted, and caused the articular surfacesto become increasingly transparent. A number of cupsdisplayed fine, radial scratches but very few cups showeddeep scratches (Fig. 5(d)).The natural joint operates under a full fluid film

lubrication regime, where the load is carried by thesynovial fluid, resulting in minimal contact between thearticulating surfaces. This favourable situation leads toa great reduction in wear in the articulation. Ascompliant layers suffer considerable friction damage inconditions of poor lubrication [36], some frictiondamage would be expected in the absence of an effectivelubrication regime. Little damage of this type wasobserved however, suggesting that the low modulusCL simulates the lubrication regime of articularcartilage in the natural synovial joint. Third body wear,resulting from free particles trapped between thearticulating surfaces, can cause severe damage to boththe acetabular and femoral components. The absence ofthird body particles between the articulating surfaces, oras inclusions in the surface of the PU CL, providesfurther evidence for the presence of an effective fluid-film-lubricating layer.

larged views of the high load-bearing surface. Impingement damage in

n the shell (blue arrows).

ARTICLE IN PRESS

Fig. 5. Macroscopic features observed in the retrieved CL cups.

I. Khan et al. / Biomaterials 26 (2005) 633–643 637

3.2. Microscopic observations

Environmental scanning electron microscopy (Fig. 6),revealed no significant damage, other than burnishingand minor third-body wear on the articular surfaces ofmost retrieved cups. Asperities, apparent on unusedcups, were still present over large areas of the surface,particularly in low load-bearing regions, giving furtherevidence of fluid-film lubrication. In high load-bearingregions the asperities were flattened and worn.Fig. 6 (a)–(c) shows the effects of burnishing on

surface asperities on moving from low (a) to high load-bearing regions (c) where the networks of rounded hill-like asperities, become flattened. Three types of thirdbody abrasive damage occurred: (e) long fine scratches

orientated radially around the cup surface; (f) radiallyorientated fine scratches with a low aspect ratio; (g)relatively larger gouge features, significantly wider anddeeper than the other two types of superficial scratchmorphologies. (h) Small, localized regions of white non-calcific deposits, possibly fatty tissue deposited duringretrieval.

3.3. Surface roughness

WYKO scans for each core sample were taken overtwo 4600 mm2 areas, measuring four surface roughnessparameters (mean surface roughness, Ra; root meansquared roughness, Rq; mean of the five greatestpeak-to-valley separations, Rz; mean peak-to-valley

ARTICLE IN PRESS

Fig. 6. Environmental scanning electron micrographs of various types of surface features and damage observed in the retrieved CL cups: (a) surface

asperities; (b) smooth asperities; (c) burnished regions; (d) fine long scratches; (e) fine short scratches; (f) larger gouge feature; (g) pores and (h)

surface deposits.

I. Khan et al. / Biomaterials 26 (2005) 633–643638

separations over the entire measured area, Rt). Thevalue of Ra approximately halved over 3 years with theother roughness parameters showing a steady decreasewith implantation time.The plot (Fig. 7) of the control surface shows that,

initially, the surface asperities range in height from tensof nanometres up to 300 nm, with a positively skewedroughness (projections rather than depressions). Withincreased implantation time flattening and wear occurs,producing a highly burnished region in the 3-year cup.However the burnished region retains asperities, sug-gesting no significant wear-mediated loss of material.

3.4. Microstructure

DSC analysis revealed a decrease in soft segment glasstransition temperature (Tg) with increasing implantation

time. The significantly lower value apparent for the3-year cup suggested increased phase separation. Nosignificant difference in Tg between the dry and wetcontrols was observed, suggesting that the additionalfactor of articulation-induced stress in the CL may havecaused an increase in phase separation [1].The melting endotherms were labelled T1, T2, T3, and

T4. Significantly higher T1 temperatures were measuredin the retrieved cups than in the dry controls, andslightly higher temperatures than the wet controls,suggesting that the refinement of this short orderendotherm may also be slightly influenced by stressesin the CL. The second endotherm, T2, occurred between145–150�C, while the third endotherm, T3, occurred inthe temperature range 156–160�C. A higher temperatureendotherm, T4, (184–189

�C) was observed for the 2-and3-year wet control and retrieved samples along with

ARTICLE IN PRESS

Fig. 7. 3D plots of typical articular surfaces from Region 1 of the retrieved cups. The control shown is a 2 year PBS-incubated sample. Scanned area:

75.4� 56.5mm=4600mm2.

I. Khan et al. / Biomaterials 26 (2005) 633–643 639

additional tiny melting peaks, at higher temperatures inthe retrieved samples than in the control cups. Thetemperatures of the endotherms measured for region 1of the 2- and 3-year cups were higher than region 2 withthe highest at 235�C (3-year cup), possibly resultingfrom microcrystalline melting [37–39], further evidencefor stress-induced crystallisation producing more or-dered microcrystalline structures.Furthermore, enzymes in body fluids have shown a

propensity to hydrolize PCUs [40–45]. Santerre et al.demonstrated that PCUs with a higher hard segmentcontent had a greater resistance to hydrolytic degrada-tion, and that enzyme-catalyzed degradation of poly-carbonate-based urethanes was initiated in amorphoussoft segment regions close to the interface between hardand soft segments. A higher hard segment content at thesurface contributed significantly to an increase inbiostability [46–48]. These results show that incubationin an aqueous environment at 37�C (whether in vitro orin the in vivo synovial joint), resulted in an improvementin the short-range order of the hard microdomain [1,35].

3.5. Mechanical properties

Load-penetration tests at three different depthsassessed changes in the macro-stiffness of the articularlayer. The loads required to produce displacements of0.14 (tenth-depth), 0.35 (quarter-depth) and 0.70mm(half-depth) were used to indicate changes in stiffness,with the load taken as being proportional to the elasticmodulus (E).

Samples from region 1 (Fig. 8(a)) showed significantlyhigher load values at all three displacements, especiallythe 1 year and 2 year cups, whilst those from region 2(Fig. 8(b)) showed no significant differences in load atthe three displacements. This implies an increase in E inthe high load-bearing region 1, but no significant changein the low load-bearing region 2, consistent with theDSC observations of improved hard microdomainstructures. The absence of any significant decrease inpenetration load (stiffness) with implantation time forboth regions suggested that there was no significantCorethane 80A biodegradation.Nano-indentation was used to investigate micro-scale

changes in the hardness and elastic modulus of thearticular surface (approximately the upper 2 mm) of thecores. Using a NanoTest 600 nano-indenter, fiveindents, 50 mm apart, under a 0.5mN maximum loadwere made in samples from region 1. Hardness and E

were determined using the Oliver and Pharr method [49].There was a significant increase in hardness and E forthe samples from the retrieved cups compared withcontrols confirming previous observations, with littlesignificant difference between the 1-, 2- and 3-year cups.

3.6. Chemical properties

Infra-red spectra were obtained for both regions ofinterest using the 1591 cm�1 band corresponding to thearomatic n(CQC) in the MDI of the hard segment asan internal reference peak, since its position is un-affected by in vivo implantation [50].

ARTICLE IN PRESS

Fig. 8. Loads at 0.14, 0.35 and 0.70mm penetration depths for regions 1(a) and 2(b).

I. Khan et al. / Biomaterials 26 (2005) 633–643640

The polycarbonate soft segment of Corethane 80Ahas major absorption bands at 1737 cm�1 (n(CQO)carbonate carbonyl with a slight contribution from theurethane carbonyl) and at around 1250 cm�1 (n (C–O–C)in (CO–O–C)––actually detected at 1243 cm�1) [51].No significant difference was observed between thespectra of the control and retrieved cups (Fig. 9).Likewise, no significant difference between the controland retrieved cups was observed in the 1700 cm�1 band(hydrogen-bonded urethane CQO) of the hard segment(Fig. 9).

3.7. Molecular weight changes

None of the samples under test, which were all fromthe high load-bearing region, showed a significantdecrease in molecular weight following incubation inair or PBS, or after implantation. The most abundantsample set (2-year, 10 samples) showed the largestdecrease (15%). Three of the samples in this set had

slightly lower molecular weights than the 2-year control,while a further three were significantly lower (andgreater than 15%). Three samples had slightly highermolecular weights. The 3-year cup was in the sameregion as the control samples.Losses were attributed to chain scission, while gains

were attributed to cross-linking. However, taken as awhole, none of the changes were significant (Table 1).

4. Conclusions

The real-time and accelerated in vitro tests reported inPart I examined the main degradation mechanisms(hydrolysis, ESC, MIO and calcification) observed inbiomedical PUs [1]. Of the four materials tested,Corethane 80A, was shown to be most suited to long-term implantation. The PUs with a polycarbonate softsegment (Corethane 80A and ChronoFlex 80A) were themore biostable materials. The work on in vivo retrieved

ARTICLE IN PRESS

16501670169017101730175017701790

Wavenumber (cm-1)

Abs

orba

nce

Control

3yr

2yr

1yr

Fig. 9. Change in n(CQO) (1737 cm�1 and 1700 cm�1) absorbances with implantation time for the retrieved cups (region 1). The control is a 2 year

PBS-incubated sample.

Table 1

Molecular weight averages (Mw), molecular number averages (Mn),

and polydispersity values (Mw/Mn) for region 1 of the control and

retrieved CL cups

Sample Mw Mn Mw/Mn

1 year dry control 80 1007707 34 0507212 2.470.04

1 year wet control 89 3507778 42 80071 414 2.470.08

1 year retrieval 77 967710 073 34 25073 447 2.370.11

2 year dry Control 76 40071 414 32 40071 131 2.470.03

2 year wet Control 76 1007283 35 2007283 2.270.09

2 year retrieval 65 635712 162 30 80075 049 2.170.12

3 year dry Control 77 45071 485 33 45071 202 2.370.05

3 year wet Control 75 9507212 35 150771 2.270.02

3 year retrieval 74 65071 344 34 150771 2.270.04

I. Khan et al. / Biomaterials 26 (2005) 633–643 641

samples of Corethane 80A reported here validate thein vitro studies, undertaken with Corethane 80A.The absence of significant physical or chemical

degradation suggests that the prototype CL acetabularcups functioned well in the sheep THA model during a3-year period of study, with the compliant layer showingexcellent resistance to biodegradation. Despite thenumerous variables that affect each implant, there wasremarkable consistency in the condition of the cupsretrieved from the time series study.Incubation in an aqueous environment at 37�C

(whether in vitro or in the in vivo synovial joint)improved the short-range order of the hard microdo-main, which in turn may increase the modulus of thePU. In-service stresses may also improve short-rangeorder.Although the primary aim of this study was to

evaluate changes in the physicochemical properties of

the softer, more susceptible PU grades (Shore durometerhardness 80A–85A), physical damage to the inherentlybiostable (high hard segment content) Corethane 75Dshell was also evaluated. In only one example, did theCorethane 75D shell fracture. The cause could notunambiguously be assigned, but may have beenassociated with impingement and fatigue. However,shell fracture however did not result in cup loosening orin any deterioration in the Corethane 80A soft-layermaterial, which remained intact [52].Impingement wear on the cup hood and shell was a

frequent finding and may need to be addressed. Theabsence of other significant wear confirms the clinicalfindings, which showed no significant soft tissuedamage, wear particle generation or inflammatory tissueresponse [31].The lack of third-body wear damage on most of the

retrieved cups, and the absence of third body particles inthe articular surface, confirm laboratory studies suggest-ing that the CL cups can mimic the natural synovialjoint by producing, promoting and preserving aneffective lubrication regime. Where it was present, theCL cups were insensitive to even severe third bodydamage. The excellent condition of the articular surfacesof the CL joints and the results from the materialevaluation confirm that there was good joint functionover the implant period [31].The findings in this study provide compelling evidence

for the biostability and efficacy of Corethane 80A as abearing surface for compliant layer cups in this specificimplant site, and concur with the findings from thetribological and histological studies [31], by concludingthat there are no contra-indications against humanclinical trials.

ARTICLE IN PRESSI. Khan et al. / Biomaterials 26 (2005) 633–643642

Acknowledgements

This work was supported by a CASE Award from theEPSRC and Stryker Howmedica Osteonics.

References

[1] Khan I, Cameron RE, Smith N. Analysis and evaluation of a

biomedical polycarbonate urethane tested in an in vitro study and

an ovine arthroplasty model. Part I: material selection and

evaluation. (Biomaterials, in press).

[2] Okumura Y, Omori H, Ooki HH, Baba MO. The effects of cup

abduction angle on the stability, PE wear of acetabular

component in cementless total hip arthroplasty. Transactions of

the 45th Annual Meeting, Orthopaedic Research Society, 1999.

p. 928.

[3] Malchau H, Herberts P. Revision and re-revision rate in THR: a

revision––risk study of 148 359 primary operations. Prognosis of

total hip replacement, Department of orthopaedics, University of

G .oteborg, Sweden; 1998.

[4] Alexander JW, Kamaric E, Noble PC. The accuracy of acetabular

reaming in total hip replacement. Transactions of the 45th Annual

Meeting, Orthopaedic Research Society, 1999. p. 906.

[5] Moody J, DiGioia A, Jaramaz B, Blackwell M, Colgan B, Nikou

C. Inaccuracy of mechanical guides used for acetabular compo-

nent alignment in total hip replacement. Transactions of the 45th

Annual Meeting, Orthopaedic Research Society, 1999. p. 278.

[6] Mai MT, Schmalzried TP, Dorey FJ, Campbell PA, Amstutz HC.

The contribution of frictional torque to loosening at the cement-

bone interface in Tharies hip replacements. J Bone Joint Surg Am

1996;78:505–11.

[7] Shanbhag AS, Jacobs JJ, Glant TT, Gilbert JL, Black J, Galante

JO. Composition, morphology of wear debris in failed uncemen-

ted total hip replacement. J Bone Joint Surg Br 1994;76:60–7.

[8] Manley MT, Serekian P. Wear debris: an environmental issue in

TJR. Clin Orthop 1994;298:137–46.

[9] Cooper RA, McAllister CM, Borden LS, Bauer TW. Polyethylene

debris-induced osteolysis and loosening in uncemented total hip

arthroplasty. A cause of late failure. J Arthroplasty 1992;7:

285–90.

[10] Thanner J. The acetabular component in total hip arthroplasty:

evaluation of different fixation principles. Acta Orthop Scand

Suppl 1999;286:141.

[11] Kurtz SM, Muratoglu OK, Evans M, Edidin AA. Advances in the

processing, sterilization and crosslinking of ultra high molecular

weight polyethylene for total joint arthroplasty. Biomaterials

1999;20:1659–88.

[12] McKellop H, Shen FW, Di Maio W, Lancaster JG. Wear of

gamma-crosslinked polyethylene acetabular components. Clin

Orthop 1999;369:73–82.

[13] Muratoglu OK, Bragdon CR, O’Connor DO, Skehan H, Delaney

J, Jasty M, Harris WH. The comparison of the wear behaviour of

four different types of cross-linked acetabular components.

J Orthop Res 2001;19(6):1210.

[14] Laurent MP, Yao JQ, Gilbertson LN, Swarts DF, Crowninshield

RD. Wear of highly cross-linked UHMWPE acetabular liners

under adverse conditions. Transactions of the 25th Annual

Meeting of the Society of Biomaterials, 2000. p. 874.

[15] Li S, Burstein AH. Current concepts review UHMWPE. J Bone

Joint Surg Am 1994;76:1080–90.

[16] Laurent MP, Yao JQ, Bhambri SK, Gselli RA, Gilbertson LN,

Swarts DF, Crowninshield RD. High cycle wear of highly cross-

linked UHMWPE acetabular liners evaluated in hip simulator.

Transactions of the 26th Annual Meeting of the Society of

Biomaterials, 2000. p. 851.

[17] Kellop HA, Di Maio WG, Shen F-W, LuB. Wear of gamma

radiation crosslinked polyethylene acetabularcups after ageing

and against roughened femoral heads. Transactions of the 45th

Annual Meeting, Orthopaedic Research Society, 1999. p. 76.

[18] Muratoglu OK, Bragdon CR, O’Connor DO, Jasty M, Harris

WH. A comparison of 5 different types of highly cross-linked

UHMWPEs: physical properties and wear behaviour. Transac-

tions of the 45th Annual Meeting, Orthopaedic Research Society,

1999. p. 77.

[19] McKellop H, Shen FW, Liu B, Campbell P, Solovey R.

Development of an extremely wear-resistant ultra high molecular

weight polyethylene for total hip replacements. J Orthop Res

1999;17:157–67.

[20] James SP, Lee KR, Beauregard GP, Rentfrow ED, McLaughlin

JR. Clinical wear of 63 ultra- high molecular weight poly-

ethylene acetabular components. J Biomed Mater Res 1999;4B:

374–84.

[21] Shaw JH. The effect of gamma irradiation of UHMWPE medical

devices agency––lit review; 1997. http://www.medical-devices.

gov.uk

[22] Shinde A, Salovey R. Irradiation of UHMWPE. J Polym Sci:

Polym Phys Ed 1985;23:1681.

[23] Naidu SH, Bixler BL, Moulton MJR. Radiation-induced physical

changes in UHMWPE implant components. Orthopaedics

1997;20:137–42.

[24] Bruns DP, Olmstead ML, Litsky AS. Technique and results for

total hip replacement in sheep: an experimental model. Vet Comp

Orthop Traumatol 1996;9:158–64.

[25] Brumby SA, Howie DW, Pearcy MJ, Wang AW, Nawana NS.

Radiographic and histological analysis of cemented double-

tapered femoral stems. Clin Orthop 1998;355:229–37.

[26] Bergmann G, Graichen F, Rohlmann A. Hip joint forces in sheep.

J Biomech 1999;32(8):769–77.

[27] Bergmann G, Siraky J, Rohlmann A, Koelbel R. A comparison of

hip joint forces in sheep, dog and man. J Biomech

1984;17(12):907–21.

[28] Bruns DP, Olmstead ML, Litsky AS. Technique and results for

total hip replacement in sheep: an experimental model. Vet Comp

Orthop Traumatol 1996;9:158–64.

[29] Unsworth A. Recent developments in the tribology of artificial

joints. Tribol Int. 1995;28(7):485–95.

[30] Dowson D. The lubrication of synovial joints. Proceedings of the

9th Canadian Congress of Applied Mechanics, University

Saskatchewan, Canada; 1983, p. 15–30.

[31] Dowson D, Fisher J, Jin Z-M, Auger DD, Jobbins B. Design

considerations in cushion form bearings for artificial hip joints.

Proc Inst Mech Eng 1991;205(Part H):52–68.

[32] Smith N, Doyle C, Jones E. Prosthetic bearing element and

process for making such an element European Patent No

EP698382, 1996.

[33] Field JR, Aberman H, Carbone A, Sharpe P, Smith N, Dunlop D,

Howie DW. An ovine model for total hip replacement: operative

procedure and complications. Vet Comp Orthop Traumatol

2001;14:32–9.

[34] Vernon-Roberts B. Institutional ethics committee guidelines,

University of Adelade, Australia. Personal communication; April

1997.

[35] Carbone A, McGee M, Field J, Howie AH, Smith N, Jones E.

Compliant layer sheep hip arthroplasty project: results of an

in vivo time series study. Vet J Biomater, in draft.

[36] Caravia L, Dowson D, Fisher J. Start up and steady-state friction

of thin polyurethane layers. Wear 1993;160(1):191–7.

[37] Martin DJ, Meijs GF, Gunatillake PA, Renwick GM, McCarthy

S. The effect of average soft segment length on morphology and

ARTICLE IN PRESSI. Khan et al. / Biomaterials 26 (2005) 633–643 643

properties of a series of polyurethane elastomers. 2. SAXS-DSC

annealing study. J Appl Polym Sci 1997;64:803–17.

[38] HuW, Koberstein. The effect of thermal annealing on the thermal

properties and molecular weight of a segmented poly-

urethane copolymer. J Polym Sci: Part B: Polym Phys 1994;32:

437–46.

[39] Eisenbach CD, Baumgartner M, Gunter C. Advances in

Elastomers and Rubber Elasticity. In: Lal J, Mark JE, editors.

New York: Plenum; 1986.

[40] Santerre JP, Labow RS. The effect of hard segment size on the

hydrolytic stability of polyether-urea-urethanes when exposed to

cholesterol esterase. J Biomed Mater Res 1997;36:223–32.

[41] Smith R, Oliver C, Williams DF. The enzymatic degradation of

polymers in vitro. J Biomed Mater Res 1987;21:991–1003.

[42] Labow RS, Erfle DJ, Santerre JP. Elastase-induced hydrolysis

of synthetic solid substrates: poly(ester-urea-urethane) and poly

(ether-urea-urethane). Biomaterials 1999;32(8):769–77.

[43] Santerre JP, Labow RS, Adams GA. Enzyme-biomaterial

interactions: Effect of biosystems on degradation of polyur-

ethanes. J Biomed Mater Res 1993;27:97–109.

[44] Santerre JP, Labow RS, Duguay DC, Erfle D, Adams GI.

Biodegradation evaluation of polyesterurethanes with oxi-

dative and hydrolytic enzymes. J Biomed Mater Res 1994;28:

1187–99.

[45] Labow RS, Duguay DG, Santerre JP. The enzymatic hydrolysis

of a synthetic biomembrane: a new substrate for cholesterol and

carboxyl esterases. J Biomater Sci Polym Ed 1994;6:169–79.

[46] Tang YW, Labow RS, Santerre JP. Enzyme-induced biodegrada-

tion of polycarbonate polyurethanes: dependence on hard-

segment concentration. J Biomed Mater Res 2001;56(4):516–28.

[47] Tang YW, Labow RS, Santerre JP. Enzyme-induced biodegrada-

tion of polycarbonate-polyurethanes: dependence on hard-seg-

ment chemistry. J Biomed Mater Res 2001;57(4):597–611.

[48] Tang YW, Labow RS, Revenko I, Santerre JP. Influence of

surface morphology and chemistry on the enzyme catalyzed

biodegradation of polycarbonate-urethanes. J Biomater Sci

Polym Ed 2002;13(4):463–83.

[49] Oliver WC, Pharr GM. An improved technique for determining

hardness and elastic modulus using load and displacement sensing

indentation experiments. J Mater Res 1996;7:1564–83.

[50] Wu Y, Sellitti JM, Anderson A, Hiltner. An FTIR-ATR

investigation of in vivo poly (etherurethane) degradation. J Appl

Polym Sci 1992;46:201–11.

[51] Dillon JG. Infrared spectroscopic atlas of polyurethanes. Lan-

caster, Pennsylvania: Technomic Publishing; 1989.

[52] Dowson D. Fluid film lubrication Osborne Reynolds Centenary.

Proceedings of the 13th Leeds–Lyon Symposium on Tribology,

Leeds, 1986.