http://pih.sagepub.com/Medicine

Engineers, Part H: Journal of Engineering in Proceedings of the Institution of Mechanical

http://pih.sagepub.com/content/early/2012/03/09/0954411912439824The online version of this article can be found at:

 DOI: 10.1177/0954411912439824

March 2012 published online 9Proceedings of the Institution of Mechanical Engineers, Part H: Journal of Engineering in Medicine

James M Fick and Daniel M Espinotissue health and hydration

Articular cartilage surface failure: an investigation of the rupture rate and morphology in relation to  

Published by:

http://www.sagepublications.com

On behalf of: 

  Institution of Mechanical Engineers

can be found at:Proceedings of the Institution of Mechanical Engineers, Part H: Journal of Engineering in MedicineAdditional services and information for

    

  http://pih.sagepub.com/cgi/alertsEmail Alerts:

 

http://pih.sagepub.com/subscriptionsSubscriptions:  

http://www.sagepub.com/journalsReprints.navReprints:  

http://www.sagepub.com/journalsPermissions.navPermissions:  

What is This? 

- Mar 9, 2012OnlineFirst Version of Record >>

at The University of Auckland Library on March 13, 2012pih.sagepub.comDownloaded from

Original Article

Proc IMechE Part H:J Engineering in Medicine0(0) 1–8� IMechE 2012Reprints and permissions:sagepub.co.uk/journalsPermissions.navDOI: 10.1177/0954411912439824pih.sagepub.com

Articular cartilage surface failure: aninvestigation of the rupture rate andmorphology in relation to tissue healthand hydration

James M Fick1 and Daniel M Espino2

AbstractThis study investigates the rupture rate and morphology of articular cartilage by altering the bathing environments ofhealthy and degenerate bovine cartilage. Soaking tissues in either distilled water or 1.5 M NaCl saline was performed inorder to render the tissues into a swollen or dehydrated state, respectively. Creep compression was applied using an8 mm flat-ended polished indenter that contained a central pore of 450 mm in diameter, providing a consistent region forrupture to occur across all 105 tested specimens. Rupture rates were determined by varying the nominal compressivestress and the loading time. Similar rupture rates were observed with the swollen healthy and degenerate specimens,loaded with either 6 or 7 MPa of nominal compressive stress over 11 and 13 min. The observed rupture rates for thedehydrated specimens loaded with 7 MPa over 60 and 90 s were 20% versus 40% and 20% versus 60% for healthy anddegenerate tissues, respectively. At 8 MPa of nominal compressive stress over 60 and 90 s the observed rupture rateswere 20% versus 60% and 40% versus 80% for healthy and degenerate tissues, respectively; with all dehydrated degener-ate tissues exhibiting a greater tendency to rupture (Barnard’s exact test, p \ 0.05). Rupture morphologies were onlydifferent in the swollen degenerate tissues (p \ 0.05). The mechanisms by which dehydration and swelling induce initialsurface rupture of mildly degenerate articular cartilage differ. Dehydration increases the likelihood that the surface willrupture, however, swelling alters the observed rupture morphology.

KeywordsCartilage, surface rupture, osmotic environment, soft tissue mechanics, rupture rate

Date received: 1 August 2011; accepted: 1 February 2012

Introduction

This article reports on articular cartilage surface rup-ture rates and rupture morphologies during compres-sion. Both healthy and mildly degenerate tissues undertwo different states of hydration are compared.Although tissue hydration has been previously shownto influence the level of nominal stress and loading timerequired to rupture articular cartilage1 it has not beenutilized to investigate how the rates of articular carti-lage rupture and the observed rupture morphology areaffected.

Articular cartilage is a load-bearing material, presentin joints, that is compressed during daily activities.Fluid content within articular cartilage assists in loadbearing during tissue deformation.2 Changes to fluidcontent occur during initial stages of osteoarthritiswhere fluid content increases,3 whereas cartilagebecomes dehydrated with advancing age.4 The articular

surface of cartilage is also integral to both load bearingand stress transmission.5 The process of degeneration,occurring during osteoarthritis, involves the rupturingof the surface, which decreases the above mentionedbiomechanical functions.1,5 Also affected is joint articu-lation, where an increase in friction associated withdegeneration contributes to the long term progressionof osteoarthritis.6 Thus, understanding surface ruptureis important, both for injurious compressive loading aswell as the pathology of osteoarthritis.7–11

1Most Recent Affiliation: Department of Chemical and Materials

Engineering, University of Auckland, New Zealand2School of Mechanical Engineering, University of Birmingham, UK

Corresponding author:

James M Fick, Most Recent Affiliation: Department of Chemical and

Materials Engineering, University of Auckland, Private Bag 92019,

Auckland, New Zealand.

Email: jfickbiomaterials@yahoo.com

at The University of Auckland Library on March 13, 2012pih.sagepub.comDownloaded from

Previous studies have determined how certainmechanical factors contribute to cartilage surface rup-ture. For example, both compressive stress and the levelof in-plain strain have been investigated and shown toinduce rupture.12–15 Rupture rates increased withincreasing peak stress and varied with strain-rate.16

Removing fluid from cartilage mechanically, eitherthrough pre-loading or creep compression of cartilage-on-bone specimens, reduces the likelihood of surfacerupture.17,18 These latter studies are important becausethey show that fluid content can alter the propensity ofthe cartilage surface to rupture. A recent study, there-fore, compared how large differences in tissue hydra-tion altered surface rupture.1 Dehydrated cartilagefailed at greater induced nominal compressive stress, inagreement with previous studies.17,18 However, tissuehydration had a greater influence in altering the time tofailure over the loading required to induce surface rup-ture than such stress. The time required to induce sur-face rupture varied by an order of magnitude betweenswollen and dehydrated samples.1 It was proposed thatthe mechanism by which failure was induced differedbetween dehydrated and swollen tissue. With the for-mer likely failing owing to stress concentrations inducedwithin the tissue and the latter failing owing to anincreased internal-swelling pressure present within thematrix. Differences in surface weakness were notassessed using surface rupture morphology. However,surface rupture morphology has been used previouslyto assess the extent of surface weakness with multiplerupture branches indicative of cartilage that has aweaker intrinsic surface strength.13 Such assessmentmay enable changes to surface strength, induced bychanges to tissue hydration, to be better understood.

In the previous study by Fick and Espino,1 condi-tions that guaranteed surface rupture were used.However, the transition from cartilage that does notundergo surface rupture, to that which undergoes sur-face rupture, enables investigation of initial cartilagefailure. Previously, surface rupture was determined tobe more dependent on tissue hydration than milddegradation of tissue structure.1 However, whethermildly degenerate cartilage becomes more vulnerable torupture during initial surface failure of cartilage, ratherthan under conditions that guarantee failure, is unclear.

Therefore, the aim of this study is to investigate initialstages of surface rupture (i.e. following transition fromnon-failure to failure), following large changes in tissuehydration for surface-intact healthy and mildly degener-ate tissues (i.e. a model used to study the early effects ofdegeneration). Surface rupture morphology is assessedfollowing the initial surface rupture of cartilage.

Materials and methods

Cartilage samples

A total of 105 cartilage-on-bone test specimens wereobtained from the distal-lateral quadrant of bovine

patellae, stored at –20 �C, and thawed prior to testing.Both healthy and mildly degenerate patellae were usedfor these experiments. The tissue preparation procedurehas been described previously;1 briefly, the patellaewere macroscopically assessed for superficial fibrilla-tion (i.e. surface degeneration) with Indian-ink19 usingthe Outerbridge scale.7 The most distal-lateral regionsthat contained an intact surface on all patellae wereselected. Mildly degenerate specimens that wereselected were always adjacent to the surface lesion orarea of fibrillation on the patellae (Figure 1). Test spe-cimens consisted of blocks of articular cartilage(approximately 153 15mm) with approximately10mm depth of underlying bone. Specimens were equi-librated in either distilled water (i.e. swollen cartilage)or a 1.5MNaCl saline solution (i.e. dehydrated carti-lage). Equilibrations lasted 12h (at 4 �C) to obtain amaximum/minimum internal hydration pressure.5

Healthy and mildly degenerate cartilage specimenswere tested when either swollen or dehydrated in abathing solution that corresponded to their equilibra-tion solution. After every test, the surfaces of the spe-cimens were inspected for rupture. If any specimenfrom its corresponding group ruptured, then no fur-ther tests were performed on that group of specimens.If rupture did not occur across a specimen group,then the tissues were retested. Any sample undergoingrepeat testing was re-equilibrated at 4 �C for 4 h in itsrespective solution. The rates of surface rupture forboth healthy and mildly degenerate tissues were deter-mined when:

(a) swollen or dehydrated were tested at incrementallevels of nominal compressive stress at fixed load-ing times;

Figure 1. Sample grade I cartilage specimen. Regions next tothe degenerate area were obtained from the distal-lateralportion of patellae. Scale bar ~1 cm. Such regions havepreviously been shown to undergo structural changes.20,21

2 Proc IMechE Part H: J Engineering in Medicine 0(0)

at The University of Auckland Library on March 13, 2012pih.sagepub.comDownloaded from

(b) swollen or dehydrated were tested for incrementaltime periods at fixed levels of nominal compressivestress.

This led to four loading regimes (as detailed below),these are detailed in Table 1.

Swollen test specimens loaded over fixed times. The initialstarting nominal compressive stress that was used was5MPa (arbitrarily chosen, but based on our previousrupture findings1) with a loading time of approximately15.4min. Tissue samples were tested with 5, 6, and7MPa of nominal compressive stress.

Swollen test specimens loaded with fixed levels of nominal com-pressive stress. The initial starting nominal compressivestress that was used was 6MPa, as surface rupture wasobserved in the previous experiments utilizing fixedloading times. The test specimens were loaded for7.5min, approximately half the time required to reachpeak pore pressure.5 Tissue samples were tested withfixed levels of nominal compressive stress of 6 and7MPa over 7.5, 11 and 13min (i.e. increasing towardspeak pore pressure).

Dehydrated test specimens loaded over fixed times. The ini-tial starting nominal compressive stress that was usedwas 6MPa (chosen because swollen test specimens wereobserved to have ruptured with this level of nominalcompressive stress) with a loading time of approxi-mately 15.4min.

Because the time to reach peak pore pressure wasdetermined from prior work performed on samplesthat were equilibrated in distilled water,5 the fixedloading time was verified in order to determine if rup-ture was occurring before 15.4min. In order toachieve this, three sets of healthy dehydrated tissuegroups were selected and loaded with 7MPa over30 s, 3min and 7.5minutes. Surface rupture wasobserved in the groups loaded only over 3min and7.5min, however, no ruptures were observed in thetissues loaded over 30 s. Thus, a fixed loading time-frame of 3min was selected to examine the dehy-drated tissue specimens.

Another set of dehydrated tissues were loaded witha nominal compressive stress of 6MPa over a loadingtime of 3min. Tissue samples were tested with 6, 7, and8MPa of nominal compressive stress.

Dehydrated test specimens loaded with fixed levels of nominalcompressive stress. The initial starting nominal compres-sive stress that was used was 7MPa, as surface rupturewas observed in the previous experiments utilizing fixedloading times. Tissue samples were tested with fixed lev-els of nominal compressive stress of 7 and 8MPa over0.5, 1 and 1.5min.

Indentation

Indentation was performed with a cylindrical polishedindenter (8mm in diameter) and included a central pore(450mm in diameter). This induced a consistent rupturesite. Mounting required setting specimens in plaster, ina bath solution. All samples under-went compressivecreep tests for the load/time previously defined; with alinear variable displacement transformer (LVDT) usedto measure axial displacement.1

Statistical analysis

Both the percentage of ruptured samples for each testprotocol and the observed rupture morphology wererecorded. The number of ruptures was pooled acrosseach tissue group for each testing protocol. A one-tailedBarnard’s exact test (BET) was performed (p \ 0.05)in order to compare the number of observed ruptureevents (i.e. rupture frequency) between the healthy andmildly degenerate test specimens. More informationabout the BET can be found elsewhere.22–25 Surfaceruptures for these experiments were recorded as eitherexhibiting single rupture or multiple branches. The rup-ture morphologies were pooled together for each speci-men group and Mann-Whitney statistical tests wereperformed. The null hypothesis was that there were nodifferences observed for the rupture morphologiesbetween the healthy and mildly degenerate specimensthat were either swollen or dehydrated (p \ 0.05).

Results

Rupture rates

Significantly greater rupture rates (p \ 0.05) wereobserved in the dehydrated mildly degenerate tissuesloaded over 90 s at either 7 or 8MPa when comparedwith the dehydrated healthy test specimens (20% versus60% and 40% versus 80%, respectively; Figure 2).Similar rupture rates were determined for swollen

Table 1. Number of samples used for each protocol.

Protocol (a) swollen Protocol (b) dehydrated Total

Constant stress Constant time Constant stress Constant time

Healthy 20 10 35 10 75Mildly degenerate 20 10 25 10 65Total 40 20 55 20 105

Fick and Espino 3

at The University of Auckland Library on March 13, 2012pih.sagepub.comDownloaded from

healthy and mildly degenerate tissues loaded with 6 and7MPa over 11, 13 and 15,4min of loading (see Figure3(a) and (b)). Rupture rates were also similar for thehealthy and mildly degenerate tissues loaded with 7 and8MPa over a fixed loading time of 3min (Figure 3(a)).The dehydrated healthy and mildly degenerate rupturerates loaded over 30 s with 8MPa and those loadedover 60 s with either 7 or 8MPa were also not signifi-cantly different (Figure 2).

Observed rupture morphologies

The swollen mildly degenerate specimens were observedto contain a greater number of rupture branches whencompared with the swollen healthy specimens (Figure4(a)–(d); p \ 0.05). The healthy rupture morphologiesextended in an approximate straight line orientationfrom the perimeter of the central-pore region out intothe directly loaded region. In addition to the abovemorphological description, the mildly degenerate rup-ture morphology also contained a branch on the oppo-site side of the central-pore region. No differences inthe number of rupture branches were observed betweenthe dehydrated healthy and mildly degeneratespecimens.

Discussion

Overview

Two main differences were found between healthy andmildly degenerate articular cartilage under transitionfrom non-failure to failure, i.e. initial cartilage surfacerupture, associated with initial degeneration.10,26,27

Dehydrated mildly degenerate cartilage tested with

either 7 or 8MPa of nominal compressive stress overeither 1 or 1.5min of loading time experienced higherfailure rates than dehydrated healthy cartilage.Conversely, the swollen mildly degenerate cartilage wasobserved to contain an increased number of rupturebranches relative to the swollen healthy tissues. Theincrease in the observed rupture branches is indicativeof the lower rupture strength of the mildly degeneratestrain-limiting surface layer, consistent with the bi-layerrupture theory.13 Also consistent with the explanationof the bi-layer rupture theory are the differences in thenumber of rupture branches owing to the loss of thedegenerate tissue intrinsic surface strength.13 This lossof surface strength is associated with the de-structuringof the collagen network and loss of proteoglycans thatmildly degenerate tissues experience during the earlystages of osteoarthritis.11,28,29 The loss of proteoglycans

Figure 3. (a). Graphical summary of the observed frequency ofrupture for swollen test specimens over loading times of 7.5, 11and 13 min. (b). Graphical summary of the observed frequencyof rupture for swollen and dehydrated test specimens whenloaded with 5, 6, or 7 MPa of nominal compressive stress. Mildlydegenerate specimens are referred to as ‘Degen’ anddehydrated tests specimens are referred to as ‘Dehyd’.

Figure 2. Graphical summary of the observed frequency ofrupture for dehydrated (‘Dehyd.’) test specimens over loadingtimes of 0.5, 1 and 1.5 min. Mildly degenerate specimens arereferred to as ‘Degen’.

4 Proc IMechE Part H: J Engineering in Medicine 0(0)

at The University of Auckland Library on March 13, 2012pih.sagepub.comDownloaded from

from cartilage has been shown to reduce compressiveresistance of cartilage,30,31 whereas the de-structuringof the matrix reduces the tensile strength of the col-lagen network.30,31

Dehydration increases the observed rupture rate

The dehydrated healthy tissue rupture rate (whenloaded with 7 or 8MPa of nominal compressive stress)was lower at both 1 and 1.5min of loading relative tothe observed rupture rates for the dehydrated mildlydegenerate specimens. Because dehydrated tissue con-tains a lower amount of water, the fluid phase duringthe initial process of load-carriage will experience alower degree of load partitioning within the tissue.32–34

This implies that the collagen network in dehydratedtissue experiences a greater degree of partitioned loadas discussed by Fick and Espino.1 Thus, the earlier col-lapsing of the collagen network will occur in a dehy-drated tissue relative to a hydrated matrix. The samedegree of load partitioning onto a dehydrated andmildly degenerate matrix would mean that more stressis concentrated onto portions of the network that canstill effectively function and transmit stress throughout

a wider continuum. The dehydrated collagen networkexperiences an even greater and/or earlier collapsing ofits network, relative to a hydrated matrix. Therefore,dehydration increases the likelihood that mildly degen-erate cartilage fails during a single compressive defor-mation that is capable of rupturing the tissue. However,the resulting surface rupture branches do not change inmorphology.

Swelling alters rupture morphology

Cartilage swelling (and therefore matrix swelling) isregulated by the support provided by the cartilage sur-face, surrounding matrix and underlying bone.35,36

Adams et al.17 and Morel et al.18 have demonstratedthat removing fluid from cartilage, either through pre-loading or creep compression of cartilage-on-bone spe-cimens, reduces the likelihood of surface rupture. Thisis in agreement with our results, as greater stress wasrequired to induce surface failure (for dehydrated overswollen specimens). Our degenerate swollen specimenscontained an increase in the number of rupturebranches relative to the swollen healthy specimens.Therefore, a combination of swelling and initial

Figure 4. Surface rupture morphologies observed in each tissue group: (a) healthy dehydrated response; (b) dehydrateddegenerate response; (c) healthy swollen response; (d) swollen degenerate response. Indian ink staining used to illustrate the loadedregions and extenuate the surface ruptures. Scale bars = 500 um.

Fick and Espino 5

at The University of Auckland Library on March 13, 2012pih.sagepub.comDownloaded from

cartilage degeneration increases the vulnerability of theweakened surface during initial failure of cartilage.This might be owing to swollen cartilage (e.g. at or nearpeak pore pressure5) having more fluid available torecruit a greater proportion of the cartilage surfaceduring load bearing.37 In turn, this induces initialfailure at the weakest points over a greater surfacearea (hence the branching). This method is different tothat described in the previous subsection entitled‘Dehydration increases the observed rupture rate’ (fordehydrated failure), as in this case stresses are lesslocalized as they are induced over a greater area.Previously, we demonstrated that such mildly degener-ate specimens showed no differences in load or stressfor failure compared with healthy specimens.1 In thisstudy, mildly degenerate specimens have been observedto undergo differences in surface rupture morphologywhen the tissue is swollen. This is important becausesurface fractures are associated with tissue degenera-tion.38 Therefore, swelling is likely to induce initial fail-ure of mildly degenerate cartilage by inducing anincreased number of surface ruptures.

Clinical relevance

These findings suggest that mildly degenerate tissueundergoes different mechanisms of failure owing tocompression when dehydrated and swollen. Duringdehydration, the greater indentation induced compres-sion of the mildly degenerate tissue1 causes the near-parallel orientation of the superficial layer of the col-lagen network. Hence the cartilage would experienceincreased tensile stresses relative to that of a healthy net-work.30 Dehydration might be induced owing to exces-sive exercise39–41 and cartilage is known to dehydratewith age.4 Removal of excess fluid has been shown tostiffen cartilage17 and even increase the strength of carti-lage,1,17,18 therefore, potentially having beneficial effectsin terms of its mechanical function. However, our resultsshow that excess dehydration may lead to initial failureof intact regions compromised by the degenerate pro-cess. This might be a mechanism of initial degenerationand its propagation through cartilage.

If mildly degenerate tissue is swollen, it is morelikely to lead to multiple surface ruptures if loaded tofailure. Swelling might occur following periods ofinactivity, however, during degeneration, tissue swel-ling is also known to occur.6 These surface cracksmay alter the permeability of the cartilage surface. Ithas been suggested that such cracks enable additionalwater to enter the tissue, thus contributing to furtherswelling.35 Further swelling could mean further crackgeneration, further propagating surface rupture andtissue degeneration.

Limitations

As with any study, there are a number of potential lim-itations. One potential limitation relates to scaling

effects caused by the similarity in scale between theindenter (8mm diameter) and specimen block(153 15mm block). However, microscopic studies haveshown such surface effects do not extend beyond 1mmfrom the loaded region in cartilage.5 In this study, noscaling effects are anticipated as the specimen edge was4mm away from the loading indenter site (i.e. fourtimes the necessary distance).

Another potential limitation is that the water contentof healthy and mildly degenerate tissues wasnot determined. The two bathing solutions utilized inthis current study provided two extremes: swollen tissue(i.e. hyper-hydrated) and dehydration. This enabledfundamental differences of surface rupture owing to tis-sue hydration to be determined (it was not the intentionof this study to correlate a range of hydration states tomeasurements of failure stress or time). Therefore, tis-sues were equilibrated in one of the two solutions and aconsistent protocol was employed for both the hydratedand dehydrated tissue groups.

Repeat testing of tissues may have introduced arti-facts that influenced the rupture characteristics of thetissues utilized in the study (i.e. fatigue, micro-damageor even visco-elastic changes to the tissues). However,McCormack and Mansour reported that, after loadingarticular cartilage with a nominal stress that rangedfrom 0.6–1.5MPa over 64,800 cycles, no changes in thetensile strength occurred.42 Although the levels of stressutilized in the current study were larger than those uti-lized by McCormack and Mansour, the number ofpotential repeat loadings was much less.

Previous studies have typically allowed between90min43,44 to 4 h1,45 for cartilage to recover to itsoriginal state. Furthermore, 3 h rebalancing in theappropriate solution leads to around 90% equilibra-tion.46 The current study allowed 4 h of rebalancing,thus, negligible differences owing to changes in vis-coelastic effects from the tissue rehydrating proce-dures are anticipated. However, any possible accruingdamage (from multiple testing), or visco-elasticeffects, would be expected to be consistent for healthyand mildly degenerate tissue groups, thus not alteringour conclusions.

The current study utilized static compression toinvestigate the failure mechanisms of cartilage that canbe considered non-physiologic. For example, Fulcher etal. previously investigated cartilage under physiologicaland impulse loading rates (1–92Hz) that are representa-tive of normal gait, with nominal values of stress up to1.7MPa.47 The nominal compressive stresses utilized inthe current study were three to four times greater thanthose utilized by Fulcher et al.47 in order to induce sur-face rupture. However, contact pressures may exceed8MPa at the hip during sitting48 or in the tibiofemoraljoint during squatting.49 The frequency and duration ofsuch activities are directly relevant to the induced stressand time-frames applied in this study (and may inducecartilage surface rupture).

6 Proc IMechE Part H: J Engineering in Medicine 0(0)

at The University of Auckland Library on March 13, 2012pih.sagepub.comDownloaded from

Conclusion

The mechanisms by which dehydration and swellinginduce initial surface rupture of mildly degeneratearticular cartilage differ. Dehydration simply increasesthe likelihood that the surface will rupture; whereasswelling alters the rupture morphology rather thanaltering the rate of rupture.

Funding

JM Fick was funded by the Health Research Councilof New Zealand, Arthritis New Zealand, and theMaurice and Phyllis Paykel Trust. DM Espino is sup-ported under a Marie Curie Intra-European Fellowshipfor Career Development within the 7th EuropeanCommunity Framework Programme (Programmenumber: FP7/2007-2013; under grant agreement num-ber 252278).

Acknowledgement

The authors thank the University of Auckland’sEngineering Faculty, for access to laboratory facilities.

Conflict of interest

The authors declare no conflict of interest.

References

1. Fick JM and Espino DM. Articular cartilage surface

rupture during compression: investigating the effects of

tissue hydration in relation to matrix health. J Mech

Behav Biomed Mater 2011; 4: 1311–1317.2. Linn FC and Sokoloff L. Movement and composition of

interstitial fluid of cartilage. Arthritis Rheum 1965; 8:

481–494.3. McDevit CP and Muir H. Biochemical changes in the carti-

lage of the knee in experimental and natural osteoarthritis

in the dog. J Bone Joint Surg 1976; 58(B): 94–101.4. Venn MF. Variation of chemical composition with age in

human femoral head cartilage. Ann Rheum Dis 1978; 37:

168–174.5. Fick JM, Thambyah A and Broom ND. Articular carti-

lage compression: how microstructural response influ-

ences pore pressure in relation to matrix health. Connect

Tissue Res 2010; 51: 132–149.6. Armstrong CG and Mow VC. Variations in the intrinsic

mechanical properties of human articular cartilage with

age, degeneration, and water content. J Bone Joint Surg

1982; 64: 88–94.7. Outerbridge RE. The etiology of chondromalacia patel-

lae. J Bone Joint Surg 1961; 43(B): 752–757.8. Weightman, BO, Freeman MA and Swanson SA. Fati-

gue of articular cartilage. Nature 1973; 244: 303–304.9. Bullough PG, Yawitz PS, Tafra L, et al. Topographical

variations in the morphology and biochemistry of adult

canine tibia1 plateau articular cartilage. J Ortho Res

1985; 3: l–16.10. Kerin AJ, Coleman A, Winsom MR, et al. Propagation

of surface fissures in articular cartilage in response to cyc-

lic loading in vitro. Clin Biomech 2003; 18: 960–968.

11. Martel-Pelletier J, Boileau C, Pelletier JP, et al. Cartilage

in normal and osteoarthritis conditions. Best Pract Res

Clin Rheum 2008; 22: 351–384.12. Flachsmann R, Broom ND and Hardy AE. Deformation

and rupture of the articular surface under dynamic and

static compression. J Ortho Res 2001; 19: 1131–1139.13. Flachsmann R, Kim W and Broom N. Vulnerability to

rupture of the intact articular surface with respect to age

and proximity to site of fibrillation: a dynamic and static

investigation. Connect Tissue Res 2005; 46: 159–169.14. Flachsmann R, Kistler M, Rentzios A, et al. Influence of

an initiating microsplit on the resistance to compression-

induced rupture of the articular surface. Connect Tissue

Res 2006; 47: 77–84.15. Sasazaki Y, Shore R and Seedhom BB. Deformation and

failure of cartilage in the tensile mode. J Anat 2006; 208:

681–694.16. Morel V and Quinn T M. Cartilage injury by ramp com-

pression near the gel diffusion rate. J Orthop Res 2004;

22: 145–151.17. Adams MA, Kerin AJ and Winsom MR. Sustained load-

ing increases the compressive strength of articular carti-

lage. Connect Tissue Res 1998; 39: 245–256.18. Morel V, Mercay A and Quinn TM. Prestrain decreases

cartilage susceptibility to injury by ramp compression in

vitro. Osteoarthritis Cartilage 2005; 13: 964–970.19. Meachim, G. Light microscopy of Indian ink prepara-

tions of fibrillated cartilage. Ann Rheum Dis 1972; 31:

457–464.20. Broom ND and Flachsmann R. Physical indicators of

cartilage health: the relevance of compliance, thickness,

swelling and fibrillar texture. J Anat 2003; 202: 481–494.21. Brown CP, Oloyede A, Moody HR, et al. A novel

approach to the development of benchmarking para-

meters for characterizing cartilage health. Connect Tissue

Res 2007; 48: 52–61.22. Barnard GA. A new test for 2 x 2 tables. Nature 1945;

3954: 177.23. Barnard GA. 2 x 2 tables: a note on E.S. Pearson’s paper.

Biometrika 1947; 34: 168–169.24. Mehta CR and Senchaudhuri P. Conditional versus uncon-

ditional exact tests for comparing two binomials. Cam-

bridge, MA: Cytel Software Corporation, 2003, pp.1–5.25. Ganju J. Size of clinical trials. In: Gad SC (ed) Clinical

trials handbook. New Jersey: John Wiley & Sons, 2009,

pp.913–931.26. Braman JP, Bruckner JD, Clark JM, et al. Articular carti-

lage adjacent to experimental defects is subject to atypical

strains. Clin Orthop Relat Res 2005; 430: 202–207.27. Gratz KR, Wong BL, Bae WC, et al. The effects of focal

articular defects on intra-tissue strains in the surrounding

and opposing cartilage. Biorheology 2008; 45: 193–207.28. McDevitt CA. Biochemistry of articular cartilage. Nature

of proteoglycan and collagen of articular cartilage and

their role in ageing and in osteoarthrosis. Ann Rheum Dis

1973; 32: 364–378.29. Maroudas A, Mizrahi J, Katz EP, et al. Physicochem-

ical properties and functional behaviour of normal and

osteoarthritic human cartilage. In: Kuettner KE,

Schleyerbach R and Hascall V.C, (eds) Articular carti-

lage biochemistry. New York: Raven Press, 1986,

pp.311–328.30. Broom ND and Silyn-Robers H. Collagen-collagen

versus collagen-proteoglycan interactions in the

Fick and Espino 7

at The University of Auckland Library on March 13, 2012pih.sagepub.comDownloaded from

determination of cartilage strength. Arthritis Rheum

1990; 33: 1512–1517.31. Swann AC and Seedhom BB. The stiffness of normal

articular cartilage and the predominant acting stress lev-els: implications for the aetiology of osteoarthritis. Br J

Rheum 1993; 32: 16–25.32. Wayne JS, Woo SL-Y and Kwan M. Application of the

u-p finite element method to the study of articular carti-lage. J Biomech Eng 1991; 113: 397–403.

33. Wayne JS. Load partitioning influences the mechanicalresponse of articular cartilage. Ann Biomed Eng 1995; 23:40–47.

34. Murkherjee N and Wayne J. Load sharing between solidand fluid phases in articular cartilage: I – experimentaldetermination of in situ mechanical conditions in a por-cine knee. J Biomech Eng 1998; 120: 614–619.

35. Parsons JR and Black J. Mechanical behavior of articu-

lar cartilage: quantitative changes with alteration of ionicenvironment. J Biomechanics 1979; 12: 765–773.

36. Summers GC, Merrill A, Sharif M, et al. Swelling ofarticular cartilage depends on the integrity of adjacentcartilage and bone. Biorheology 2008; 45: 365–374.

37. Li LP, Korhonen R, Livarinen J, et al. Fluid pressure dri-ven fibril reinforcement in creep and relaxation tests ofarticular cartilage. Med Eng Phys 2008; 30: 182–189.

38. Broom ND, Ngo T and Tham E. Traversing the intact/fibrillated joint surface: a biomechanical interpretation. JAnat 2006; 206: 55–67.

39. Luke AC, Stehling C, Stahl R, et al. High-field magneticresonance imaging assessment of articular cartilage beforeand after marathon running. Am J Sports Med 2010; 38:2273–2280.

40. Liess C, Lusse S, Karger N, et al. Detection of changes incartilage water content using MRI T2-mapping in vivo.Osteoarthritis Cartilage 2006; 10: 907–913.

41. Jurvelin J, Kiviranta I, Tammi M, et al. Effect of physical

exercise on indentation stiffness of articular cartilage in

the canine knee. Int J Sports Med 1986; 7: 106–110.42. McCormack T and Mansour JM. Reduction in tensile

strength of cartilage precedes surface damage under

repeated compressive loading in vitro. J Biomech 1998;

31: 55–61.43. Eckstein F, Tieschky M, Faber S, et al. Functional analy-

sis of articular cartilage deformation, recovery, and fluid

flow following dynamic exercise in vivo. Anat Embryol

1999; 200: 419–424.44. Broom ND and Meyers DB. A study of the structural

response of wet hyaline cartilage to various loading situa-

tions. Connect Tissue Res 1980; 7: 227–237.45. Venn M and Maroudas A. Chemical composition and

swelling of normal and osteoarthrotic femoral head carti-

lage. I. Chemical composition. Ann Rheum Dis 1977; 36:

121–139.46. Maroudas A and Evans H. Sulphate diffusion and incor-

poration into human articular cartilage. Acta biochim bio-

phys 1974; 338: 265–279.47. Fulcher GR, Hukins DWL and Shepherd DET. Viscoe-

lastic properties of bovine articular cartilage attached to

subchondral bone at high frequencies. BMC Musculoske-

letal Disorders 2009; 10: 61.48. Yoshida H, Faust A, Wilckens J, et al. Three-dimensional

dynamic hip contact area and pressure distribution dur-

ing activities of daily living. J Biomech 2006; 39: 1996–

2004.49. Wallace DA, Salem GJ, Salinas R, et al. Patellofemoral

joint kinetics while squatting with and without an exter-

nal load. J Orthop Sports Phys Ther 2002; 32: 141–148.

8 Proc IMechE Part H: J Engineering in Medicine 0(0)

at The University of Auckland Library on March 13, 2012pih.sagepub.comDownloaded from