BlOMECHANlCS OF THE
ACETABULAR LABRUM
by
Stephen John Ferguson
A thesis submitted to the Department of Mechanical Engineering
in conformity with the requirements for
the degree of Doctor of Philosophy
Queen's University
Kingston, Ontario, Canada
March, 2000
copyright Q Stephen John Ferguson, 2000 National Library Bibliothèque nationale 1+1 dca"ada du Canada Acquisitions and Acquisitions et Bibliographie Services services bibliographiques 395 Wellington Street 395. rue Wellington Ottawa ON KIA ON4 Ottawa ON K1A ON4 Canada Canada YOM fi& Votre réference
Our lile Notre rdfdmee
The author has granted a non- L'auteur a accordé une licence non exclusive licence allowing the exclusive permettant à la National Library of Canada to Bibliothèque nationale du Canada de reproduce, loan, distribute or sell reproduire, prêter, dishibuer ou copies of this thesis in microfom, vendre des copies de cette thèse sous paper or electronic formats. la forme de microfiche/nlm, de reproduction sur papier ou sur format électronique.
The author retains ownership of the L'auteur conserve la propriété du copyright in this thesis. Neither the droit d'auteur qui protège cette thèse. thesis nor substantial extracts kom it Ni la thèse ni des extraits substantiels may be printed or othenvise de celle-ci ne doivent être imprimés reproduced without the author' s ou autrement reproduits sans son permission. autorisation. Biomechanics of the Acetabular Labrum
Abstract
The goal of this research was to determine the fünctional role of the acetabular Iabrum in
the normal hip joint, and its possibIe role in the development of osreoarthrosis. Despite clinical
evidence of a Iink between labrum patholom and osteoarthrosis, there have been few studies of
the function of the acetabular Iabrurn.
An investigation of the tensile and compressive material properties of labrum tissue
showed that the labrum, with its highly onented collagen fibre structure, was much stiffer and
stronger than the adjoining cartiIage- The resistance to fluid flow through the labrum was also
higher than through cartilage. One can infer that the strength and imperrneability of the labrum
enhance its ability to seal and stabilise the hip joint.
Poroelastic finite-element models of the hip joint demonstrated that the labrum could seal
a presswised fluid layer within the hip joint under physiologica1 loading. Consequently, cartilage stresses and contact pressures were reduced. The models also indicated that, with its Iow perrneability, the labrum added an important resistance to the flou. path for fluid expressed fiom the cartiIage layers. Cartilage stresses and strains calculated by the mode1 were up to 30% higher
following removal of the labrurn. Contact pressures, and hence fnction between the cartilage surfaces, were also signilicantly higher foilowing fabrum removal.
The predictions of the finite-eIement models were evaluated through a series of in vitro whole-joint creep consolidation experiments on hurnan hips. The overall compression of the cartitage layers under a variety ofstatic and dynamic Ioads was measured. Removal of the labrum resulted in a quicker cartilage consolidation rate. Peak intra-articular fluid pressures of over 500 Wa were measured during loading in joints with a well-forrned labrum. The results OF the experiments agreed with the predictions of the finite element models and lend further support to hypotheses about labrum sealing.
S J. Ferguson Queen's University Biornechanics of the Acetabular Labrum
. - -- -
Co-Authorship
This thesis represents the results of the candidate 's independent investigation, with exceptions as noted. The technical assistance of sthers has been acknowledged in the individual chapters, as necessary. Each chapter of this thesis is the original work of the candidate. The CO-authorsof the manuscripts corntained in this thesis, Dr. J.T. Bryant, Dr.
K. Ito and Dr. R. Ganz, provided supervision of the individual studies, as well as advice with respect to the analysis of study results and manuscript revision.
S J . Ferguson iii Queen's University Biomechanics of the Acetabular Labrum
Acknowledgements
Several years ago in Toronto, as I sat designing gas-tempering loops for steel- rnaking plants, 1 would never have imagined the path ahead, which would eventually lead me to the rarefied air of the Swiss Alps. On a typical Ontario February morning I visited
Queen's University to consider returning to school for pduate studies. 1 followed Dr.
Tim Bryant al1 over the campus, running to keep up with this animated individual on his seerningly never-ending agenda. Tim's enthusiasm for his work made his classes on soft tissue biomechanics a pleasure, and when the time came to find a supewisor for this doctoral thesis, the choice for me was obvious. My introduction to the A0 Institute in
Davos was through a fortunate choice of location for a research intemship. Towards the end of this internship, 1 met Dr. Keita ho, and together we began to draw an outhe of the work presented here. We also began a fi-iendship that extends beyond the doors of the
Institute. I could not have hoped for two better supervisors- Their objective evaluation of my work throughout the project has been invaluable, although having an ocean between the two institutions has presented some logistical challenges. Thank you to both for supporting my desire to wnte manuscripts as the project progressed.
I would also like to thank Dr. ReinhoId Ganz for providing the inspiration for this research through his clinical investigations of labrum pathology, and also Dr. Slobodan
Tepic for his expert advice during a particularly hectic planning session one blustery day in Boston, and whenever my head started to hurt thinking about cartilage mechanics,
For technical assistance at the A0 Institute I am indebted to Benno Dicht and
Peter Ambühl in the machine shop, Iris KeIler and Kati Kampf for histological preparation, Emir Schluep for photography and Gethin Owen for electron microscopy.
S.J. Fermouson Queen's University Biornechanics of the Acetabular Labrum
My time at the A0 Institute and at QueenysUniversity has been made much more enjoyable by the rnany people whose paths 1have crossed. Too many good fiiends to name individual 1y; thank you al1 for the c ycling adventures, winter tours, Wok-In curries,
Schtvyzertuutsch lessons and the occasional Weizenbier in sorne cosy pub or another.
Three individuals deserve special mention. Geoff Richards provided a hearty rnkture of mischief and humour, and tireIessl y pursued the establishment of a student community.
ZouZou Kuzyk kept me sane through my comprehensive exams and I have truly missed
Our conversations since leaving Kingston. With Nick Bishop 1 have shared an office, a mountaintop, an apartment and, most importantly, a Iasting friendship. It will be a profound change to lose the daily banter and brainstorming- I wish hirn al1 the best with his owresearch, and with his devilish contraption of pipes, bearïngs and cornputers.
I have been blessed with the support of two families throughout my studies. To my mother, father and sisters 1 owe a debt of +ptitude for al1 the helpful words when 1 hit the lows and the praise when things went rny way, No matter where you go in life, you never leave home. Ich rnochte auch meiner '-Adoptivw-familie in Davos, den Ettingers, herzlich danken. Ohne ihre Unterstutzung ware ich Iangsam verrückt geworden. Danke auch, dass ihr immer einen PIatz fur mich am Tisch hattet.
Finally. to Ladina, thank you for the patience that you showed when the lab stole our time, the encouragement that you gave me to always look ahead, and the confidence that you expressed in my abilities. Every man needs sorneone special, and I'm fortunate to have found that person.
This research was conducted at the A0 ASiF Research hstitute, Davos, Switzerland and in the Department of Mechanical Engineering, Queen's University, Kingston, Ontario, Canada. Funding was provided in part by a NaturaI Sciences and Engineering Research Council (Canada) Post Graduate Scholarship and by an A0Research Foundation Grant,
SJ. Ferguson Queen's University Biomechanics of the Acetabular Labrum
for Jeianette Srone
S.J. Ferguson Queen's University Biomechanics of the Acetabular Labrum
Table of Contents
.. Abstract ...... ii ... Co-Authorship ...... iir Acknowledgements ...... iv .. Contents ...... vrl List of Tables ...... x List of Figures and Illustrations ...... x
Chapter One .Introduction ...... 1 Hypotheses ...... 5 Re ferences ...... 8 Chapter Two .Background ...... *...... *...... 10 Labrum Anatomy ...... 10 Clinical Relevance ...... 18 Etiology of Osteoarthrosis ...... ~...... 20 The Mechanical Nature of Hydrated Tissues ...... 22 Cartilage Failure Mechanisrns and Joint Lubrication ...... 25 Biomechanics of Synovial Joints ...... 29 References ...... 36
Chapter Three .3D Reconstruction ...... 45 Abstract ...... 46 Introduction ...... 47 Materials and Methods ...... 50 Dimensional Changes ...... 50 Surface Extraction ...... 54 Results ...... 55 Discussion ...... 59 Acknowledgements ...... 63 References ...... 64 C hapter Four .Material Properties ...... 66 Abstract ...... 67 Introduction ...... 68 Materials and Methods ...... 70 Specimen Preparation ...... 70 Compression Specimens ...... 71
S.J. Ferguson Queen's University Biomechanics of the Acetabular Labrum
Tension Specimens ...... 72 Dimensional Measurements ...... 73 Mechanical Testing...... 74 Con fined Compression Testing...... 74 Uniaxial Tension Testing ...... 78 Water Content ...... 82 Results ...... 83 Compressive Material Properties...... 83 Tensile Material Properties ...... 85 Discussion ...... 88 Study Design ...... 88 Compression Testing ...... 90 Tension Testing ...... 91 Acknowledgements ...... 94 References ...... 95 Chapter Five .The Labrum Seal ...... 98 Abstract ...... 99 Introduction ...... 100 Method ...... 102 Results ...... 105 Discussion ...... 110 References ...... 115 Chapter Six .Hip Joint Consolidation .Analytical ...... 118 Abstract ...... 119 Introduction ...... 120 Method ...... 123 Results ...... 127 Discussion ...... 139 References ...... 144 Chapter Seven .Hip Joint Consolidation .Experimental ...... 150 Abstract ...... 151 Introduction ...... 152 Materiais and Methods ...... 155 Specimen Preparation ...... 155 Testing Procedure ...... 158 Data Analysis ...... 161 Results ...... 163 Pressure Measurements ...... -..-...... 163 Displacement Measurements ...... 167 ...... Discussion ...... 173 References ...... 182
S.J. Ferguson viii Queen's University Biomechanics of the Acetabular Labnim
C hap ter Eight .General Discussion ...... 185 Anatomy Study ...... 186 Material Roperties ...... 190 Labrum Function ...... 191 Finite-Element Models ...... 193 In vihO Consolidation Experiments ...... 198 C linical Significance ...... 201 References ...... 203 Chapter Nine .Conclusion ...... 205 Summary ...... 205 Future Work ...... 208 Appendices ...... 212 Quantitative Anatomy ...... 212 Mechanical Stability ...... 217 Poroelastic FE Evaluation ...... 22 L Finite Elernent Mode1 Details ...... 225 Consolidation Measurement ...... 228 Vita
S.J. Fer,= on Queen's University Biomechanics of the Acetabuiar Labmrn
List of Tables
Table 4.1 Compressive Properties of the Labrum ...... 84 Table 4.2 Measured Tensile Properties of the Labrum ...... 86 Table 4.3 Denved Tensile Properties of the Labrum ...... 87 Table 6-1 Mode1 ResuIts .Pressure and Contact Stress ...... 130 Table 6.2 Mode1 ResuIts - Subsurface Strain and Stress ...... 138 Table 7.1 Consolidation Displacement Measurements ...... 171 Table 7.2 Cornpanson of Pressure Decay and ConsoIidation Rate ...... 172 Table 7.3 Cornparison of Pressure and Consolidation Time Constants ...... 173 Table E. 1 Test Configurations Simulated ...... 236
List of Figures and Illustrations
Figure 1.1 Media1 view of lefi hemipelvis ...... -3 Figure 2.1 SEM .labrum cross section ...... 12 Figure 2.2 SEM .labrum/bone interface ...... 12 Figure 2.3 SEM - labrudbone interface ...... 13 Figure 2.4 SEM .cartilagehone image ...... 14 Figure 3.5 SEM - cartiiagehone interface ...... 14 Figure 2.6 SEM - cartiIage/labrum interface ...... 15 Figure 2.7 SEM .cartiIage/labrum interface ...... 15 Figure 2.8 SEM - cartilage/labrum interface ...... 16 Figure 2.9 SEM .acetabular nm cross-section ...... 16 Figure 2-10 Cartilage stress relaxation ...... 24 Figure 2.1 1 Weeping lubrication ...... 28 Figure 2 .12 Peripheral portion of hip joint under load ...... ,...... 33 Figure 3.1 Cornparison of sectioning techniques ...... 48 Figure 3.2 Specimens for embedding ...... *...... *. 51 Figure 3.3 Dimension rneasurements ...... 53 Figure 3 -4 In-plane dimensional changes ...... 56 Figure 3.5 In-plane dimensional changes ...... 56 Figure 3.6 Stained section of acetabdar rim ...... 58 Figure 3.7 Manual definition of contour vertices ...... 58 Figure 3.8 Final 3D reconstruction ...... ,...... 59 Figure 4.1 Media1 view of bovine acetabulum ...... 68 Figure 4.2 Compression specimens ...... 72
S.J. Ferguson Queen's University Biomechanics of the Acetabular Labrum
Figure 4.3 Tension specimens ...... 73 Figure 4.4 Confined compression testinp apparatus ...... 76 Figure 4.5 Tensile testing apparatus ...... 79 Figure 4.6 Tensile testing curve fit parameters ...... 82 Figure 4.7 Compressive creep consolidation curve ...... 85 Figure 4.8 Tensile stress-strain curves ...... 86 Figure 5.1 Peripheral portion of hip joint under load ...... IO1 Figure 5.2 Axisymmetric finite element mesh ...... 103 Figure 5.33 C ircumferential strain in the Iabrum ...... 106 Figure 5.3b Circumferential stress in the labrum ...... 106 Figure 5.4a Interstitial fluid pressure (sealed)...... 108 Figure 5.4b Cartilage layer solid stresses (sealed) ...... 108 Figure 5Sa Interstitial fluid pressure (no sealins) ...... 109 Figure 5.5b Cartilase layer solid stresses (no sealing) ...... 109 Figure 6.1 The acetabular labmrn ...... 121 Figure 6.2 Plane-strain finite element mesh ...... 124 Figure 6.3a Vertical cartilage layer consolidation ...... 128 Figure 6.3b Lateral cartilage layer consolidation ...... 128 Figure 6.4 Total contact pressure (t = 1000s)...... 130 Figure 6.5a Solid-solid contact stress (t = 1000s)...... 132 Figure 6.5b Solid-solid contact stress (t = 10000s)...... 132 Figure 6.6 Total contact pressure (t = 10000s)...... L33 Figure 6.7a Principal strain (with labrum) ...... 134 Figure 6.7% Principal strain (without labrum) ...... 134 Figure 6.8a VonMises equivalent stress (with labrum) ...... 135 Figure 6.8b VonMises equivalent stress (without labrum) ...... 135 Figure 6.9a Tresca "stress" (with Iabrum) ...... 136 Figure 6.9b Tresca "stress"(without labrum) ...... 136 Figure 6.10a Principal stress (with labrum) ...... 137 Figure 6 .lob Principal stress (without labrurn) ...... 137 Figure 7.1 Whole pelvis sectionhg ...... 156 Figure 7.2 Aligment fiame ...... 156 Figure 7.3 Placement of pressure transducer...... 157 Figure 7.4 Testing apparatus in MTS servo-hydraulic frame...... 159 Figure 7.5 Fluid Pressure (static load - hip 058R)...... 164 Figure 7.6 Fluid Pressure (dynamic load - hip 058R) ...... 164 Figure 7.7 Fluid Pressure (static load - hip 170R)...... 165 Figure 7.8 Fluid Pressure (dynamic load - hip 170R) ...... 165 Figure 7.9 Fluid Pressure (static load - hip 170L)...... 166 Figure 7.1 0 Fluid Pressure (dynarnic load - hip 170L) ...... 166 Figure 7.1 1 Consolidation displacernent (static load - hip 058R) ...... 169 Figure 7.12 Consolidation displacement (dynamic Ioad - hip 058R) ...... 169 Figure 7.13 Consolidation displacement (static load - hip 227L) ...... 170 Figure 7.14 Consolidation displacement (dynamic load - hip 227L) ...... 170
S.J. Fereauson Queen's University Biomechanics of the Acetabular Labrurn
Figure 7.15 Actual pressure transducer placement ...... 175 Figure 7.16 Fissure at labnim/cartila~eintefice ...... 180 Figure 8.1 Labrum anatomy .schematic ...... 189 Figure 8.2 Labrum seal of joint space .schematic ...... 194 Figure 8.3 Additional resistance to fluid expression - schematic ...... 196 Figure 8.4 Experimental consolidation and hypotheses ...... 200 Figure A .1 Sectionins polyrner embedded hip ...... 2 13 Fipre A.2 Stained section through hip joint ...... 214 Figure A-3 Mid-coronal section of hip joint ...... 215 Figue A.4 Boundary contours extracted from serial sections ...... 216 Figure B-1 Axisymmetric mesh and pull-out displacement of femur ...... 218 Figure B.2 Pull-out force for femur ...... 219 Figure C. 1 Confined compression consolidation test ...... 222 Figure C.2 Stress relaxation test ...... 223 Figure D- 1 Fluid Iayer finite elements: open ...... 226 Figure D.2 Fluid layer finite elements: closed ...... 227 Figure E .1 Video marker clusters placement ...... 231 Figure E.2 Relative motion of marker clusters during joint consolidation ...... 231 Figure E.3 Quality of spherical fit of computer data ...... 237 Figure E.4 Absolute errors of position vectors for cornputer data ...... 237 Figure ES Maximum errors in calculated consolidation (cornputer) ...... 238 Figure E.6 Average emrs in calculated consolidation. . (compter) ...... 238 Figure E.7 Error in individual marker position ...... 239 Figure E.8 Error in calculated consolidation (bench-top) ..... ,...... 241 Figure E.9 Lndividual marker displacement for in vitro test ...... 242 Figure E .10 Cartilage layer consolidation for in vitro tests ...... 243
S.JI Ferguson Queen's University Chapter One
Introduction
Osteoarthrosis of the hip remains a common cause ofdisability in our society
[14]. Treatrnents of osteoarthrosis in the hip include non-surgical management, with
analgesics and anti-inflamrnatory dmgs, and in the case of severely degraded joint
fùnction, joint surgery such as realignment osteotomies [14,16,20] or total joint
replacement with prosthetic devices. The causes of osteoarthrosis in the hip are not fully
undeetood, but it has been proposed that changes in the biomechanical loading of the joint, either due to trauma, pre-existing joint deformities or developing joint disease. or
changes in the mechanical properties or metabolic processes of the articular cartilage, cm
lead to osteoarthrosis [3,15,18,19]. While surgical treatment with total hip replacement
enjoys a high success rate, it rvould be preferable, both financially and for the longevity of the treatment's success, to manage the problem of osteoarthrosis in the hip throu& early detection and prevention. To achieve this, it is necessary to determine the functional roles of the various anatomical structures in the hip joint in order to identiQ joint conditions that are precursors of the disease, and to better understand the etiology of osteoarthrosis,
In the last two decades, hip dysplasia in association with acetabular Iabmm injury has been a topic of investigation in the orthopaedic community. The acetabular labrurn is a fibrocartilaginous lip attached to the bony margin of the acetabular socket. It extends the coverage of the femoral head and deepens the acetabular socket. Biomechanics of the Acetabular Labrum
iliac
Figure 1. 1: Media1 view ofa lefi hemipelvis. The acetabular Iabrum (AL) encircles the anterior, superior and posterior regions of the bony acetabdar rim. lt joins smoothly with the transverse ace8tabuIar ligament (TAL) to bridge the acetabular notch, forming a complete circle.
It joins smoothly with the transverse acetabular ligament, bridging the acetabular notch and forming a complete circle (Figure 1.1). It is triangular in cross-section, with its base attached to the acetabuhm and its apex forming the fiee edge of the iabrum, which is turned in against the fernorat head [2].There is evidence that the acetabular labrum plays
S-J. Ferguson Queen's University Biomechanics of the Acetabular Labnun an important role in the development of the acetabular roof in iniàncy, and that damage to the Iabrum rnay alter the load transfer in the hip joint to such an extent that dysplasia rnay be induced in a growing hip [7,1 Il. Premature onset of osteoarthrosis (loss ofjoint space, increased subchondral-bone sclerosis, calcified osteophyte and cyst formation) has been observed in dysplastic hips treated with joint reali-ment osteotomy and requisite Iabnim removal, despite satisfactory joint reduction [6].These degenerative changes rnay have been caused by the absence of the labrum, despite the proper joint alignment. However, the relationship between hip dysplasia, acetabular labrum damage and osteoarthrosis is not clear. It is possible that damage to the labrum rnay result in the development of abnormal stresses in the hip joint, leading to bone remodehg and, eventually, hip dysplasia. Or, it is equally likely that the misalignment of a pre-existing dysplastic hip leads to abnomal shear forces across the interface between the labrum an acetabulum, causing the labrum injuries [4.12]. Tears in the labruin rnay compromise the load bearing and stability enhancing function of the labrum and, if lefi untreated, would propagate as a result of repeated stresses [9].The causal relationships of the defects are unclear and both processes rnay occur. The mechanism in either case appears to progess to the end-stage of degenerative joint disease of the hip, including OA. Labral injury alone, without accompanying hip dysplasia, rnay be suffficient to cause degenerative changes in the hip.
There have been several studies, which have shown that labral pathology, such as tears or intra-articular impingement, leads to changes in the adjoining cartilage layers consistent with early osteoarthrosis [1,4,8,10,13,17].
Although the function of the acetabular labrum has not been studied extensively, its role in the normal hip joint rnay be analogous to that of the meniscus in the knee and
S.J. Ferguson Queen's University Biomechanics ofthe Acetabular Labm
the glenoid labmm in the shoulder, due to sirnilarities in morphology or geometry. The labrum may increase the load bearing surface of the hip joint, decreasing the overall stress level in the cartilage. The acetabular labrum may also enhance joint stability, where the term stability throughout this thesis is used to express a resistance to joint subluxation. ClinicaI studies report that the rnajority of labral tears occur in the posterosuperior region of the acetabulum, a highly Ioaded region of the joint, and that these tears result in clicking and a "giving away" sensation, implying a loss ofjoint stability. A sealing function ofthe labrum has been alluded to in studies of hip joint biomechanics. The Iabrum, which is tightly apposed to the femoral head, limits the flow of fluid into and out of the intra-articular space of the hip joint. A negative pressure is developed upon luxation of the joint, and this suction effect, together with the structural resistance to subluxation offered by the labmm, is often suffkient to support the dead weight of the let [2 1,331. This is not the case following removal of the labrurn. Under compressive loading, a layer of fluid can be maintained between the opposing cartilage layers of the joint for periods of several minutes [22]. Besides improving the stability of the joint, the possible sealing knction of the labrum could improve the lubrication mechanisms in the hip joint, and also increase the loads camed by fluid pressure in the cartilage layers of the hip, thereby decreasing loading of the cartiIage solid matrix.
Failure ofthis "valve" could lead to increased fiction, higher loading in the cartilage and, eventually, to the degenerative changes of OA.
S.J. Feraouson Queen's University Biomechanics of the Acetabular Labrurn
The motivation for this research thesis is perhaps best summarised by Gardner in his review of the nature and causes of osteoarthrosis [5]:
A diarthrodial joint moves norrnally only when the actions of the skeletal muscles that act across it are physiological, when the relevant ligaments and tendons are intact, and the menisci and Iabra normal and whole. Any discussions of the origins and nature of failure of hyaline articular cartilage in osteoarthritis must therefore take into account the possible parts played by abnormalities of bone. synovial tluid, tendon, ligament, accessory shuctzrrer, and skeletal muscle.
Hypotheses
While there is clinical evidence of a possible link between labrum pathology and cartilage degeneration, the cause and effect relationship between the two is not clear.
Before attempting to predict the nature of cartilage degeneration associated with labrum patholo~y,a basic understanding of the biomechanical function of the labrurn is required.
Based on the results of the limited experimental investigation of labrum function reported to date. the following hypotheses for the biomechanical fùnction of the acetabular labrum are proposed:
1. The acetabular labrum provides a seal against intra-articular fluid flow. This seal
prevents, or at Least delays, flow of synovial fluid out of the joint space. With an
adequate seal, a pressurised fluid layer is maintained between the joint surfaces,
preventing joint contact. Hydrostatic pressure is developed in the fluid component of
the articular cartilage during loading, and the rnajority of joint forces are borne by this
fluid pressure. If the labrum is damaged, its sealing capacity is diminished. This may
S.J. Ferguson Queen's University Biomechanics of the Acetabular Labrum
allow joint surface-to-surface contact and consequently higher stresses in the solid
matrix of the cartilage.
2. The labmm extends coverage of the femoral articular surface and increases the
resistance to flow of interstitial fl uid being expressed fiom the cartilage layers of the
hip. This decreases the rate of cartilage consolidation over the course of the day
under the influence of the nominal compressive load applied to the joint. The
decreased rate of consolidation Iimits the mechanical strain, and consequently stress,
within the solid matrix of the tissue.
3. The labrurn deepens the acetabular socket and is composed predominantly of strong
collagen fibres. The labrurn is well positioned, and possesses adequate structural
strength, to bear a portion of the load applied across the joint.
The overall goal of this research is to determine the tùnctional role of the acetabular labrum in the normal hip joint. From this, one rnay inkr the possible role of the labrum in the development of joint pathologies. Topics of investigation include the sealing capacity and load canying capacity of the labrum. The main body of the thesis research has been witten in the forrn of individual manuscripts; each chapter presents a separate investigation. Where appropriate, publication and/or submission details have been provided for each manuscript.
Chapter Two provides more detailed background information on acetabular labrum anatomy, theoretical and experimental investigations of soft tissue material properties, current theones of cartilage failure mechanisrns and the etioiogy of osteoarthrosis, and experimental and analytical investigations of synovial joint biomechanics. Chapter Three
S.J. Ferwon Queen's University Biomechanics of the Acetabular Labnun
-- - -
describes a technique developed for the production of dimensionally accurate three-
dimensional reconstmctions of anatomicaI structures composed of mixed tissue types.
This technique has been applied in a quantitative anatomical study of the human hip to
provide required information about labrum size, shape and position. Chapter Four
presents an investigation into the tensile and compressive mechanical properties of
bovine labrum tissue. The results of this investigation provide some insight into the
nature of human labtwm tissue, and also provide basic parameters for the analysis of
labrum fiunction- Chapters Five and Six pertain to finite-element analysis of the capacity
of the labrum to seal the intra-articular space, and of the influence of the labruin on
cartilage Iayer consolidation in the hip joint. Chapter Seven presents the results of a
series of in vitro whole-joint creep consolidation experiments on human hips before and
after Iabrum excision. A general discussion of the results of each individual investigation
is presented in Chapter Eight to provide the reader with a cohesive link between the
vanous topics covered. Finally, the concIusions of Chapter Nine emphasise the major
findings of the thesis in the context of the clinical issues rvhich motivated the research.
As the work presented here is only a step along the way to a better understanding of the
complexities of the hip joint, directions for further research are recommended,
SJ. Ferguson Queen's University Biomechanics of the Acetabular Labrum
References
Altenberg AR: Acetabular labnim tears: a cause of hip pain and degenerative arthritis. Sorrthern :tfedÏcdJorirnal70: 174-1 75, 1977
Anonyrnous: Structure of synovial joints. In: Gr& A,mtomy., 36" edition, eds. PL Williams and P Warwick. Edinburgh, Churchill Livington, 1980
Bombelli R: Structure and F~rnctionin ~Vorrnaland Abnormal Hips: Horv ro Remte Mechanicalty Jeapordked Hip, jd edition, Berlin, Springer-Verlag, 1993
Dorrell JH, Canera11 A: The tom acetabular labrum. Joirrnal of Bone andJoint Siirgev [Br] 68-B:400-403, 1986
Gardner DL: The nature and causes of osteoarihrosis- BritAfedJ. 286d 18434, 1983
Gibson PH, Benson MK: Congenital dislocation of the hip. Review at maturity of 147 hips treated by excision of the [imbus and derotation osteotomy. Jorirnal of Bone andJoint Srtrgery [Br] 64-B: 169- 175, 1982
Graf R: [The labrum acetabulare in infants]- Orthopade 27:673-674, 1998
Han's WH. Bourne RB, Oh 1: Intra-articular acetabular labnim: a possible etioiogical tàctor in certain cases of osteoarthritis of the hip. Jorrunal of Bone andJoinr Strrgery 6 1 -A5 IO- 514,1979
Ikeda T, Awaya G, Suzuki S, Okada Y, Tada H: Tom acetabuiar labmm in young patients. Artfiroscopic diaposis and treatment. Jozrrnal of Bone andJoint Srirgery (Br] 70-B: 13- 16. 1988
ko, K.: Persona1 Communication, 1999
Kim YH: AcetabuIar dysplasia and osteoarthritis developed by the eversion of the acetabular labrum. Ciïnical Orthopaedics and Related Research 2 15:289-295, 1987
Klaue K. Durnin CW, Ganz R: The acetabular rim syndrome. A clinical presentation of dysplasia of the hip. Jotirnd of Bone and Joint Srirgery [Br] 73-B:433-429, 199 1
McCarthy JC, Busconi B: The role of hip arthroscopy in the diagnosis and treatment of hip disease. Condian Jozrrnal of Srtrgery 3 8 Suppl 1:S 13-1 7, 1995
Millis MB, Murphy SB, Poss R: Osteotomies about the hip for the prevention and treatment of osteoarthrosis. Jolirnal of Bone and Joinr Szrrgery 77-A:626-647, 1995
Mow VC, Setton LA, Guilak F, Ratcliffe A: Mechanical factors in articular cartilage and their role in osteoarthritis. In: Osteoahritic Disorders., eds, K Kuettner? V Goldberg, and IL Rosemont. Amencan Academy of Orthopaedic Surgeons, 1995
Nishina T, Saito S, Ohzono K, Shimizu N, Hosoya T, Ono K: Chiari pelvic osteotomy for osteoarthritis. Jozirnal of Bone andJoint Siirgery [Br] 72-B:765-769, 1990
S.J. Fer,ouson Queen's University Biomechanics of the Acetabular Labnun
Ohgiya H: [An arthroscopie study of coxarthrosis], ivippon Seikeigeka Gakkai Zasslzi 68:125-138, 1994
Pauwels F: Biomechanics of the Normal and DisedHip. Theoretical Formdation, Technique and Resrdts of Treatment- An Atlas,, eds. EU Furlong and P Maquet. New York, Springer, 1976
Poole AR: Imbalances of anabolism and catabolism of cartilage rnatrix components in osteoarthritis. In: Osteoarthritic Disot-ders., eds. V Kuettner, V Goldberg, and IL Rosemont, Amencan Academy of Orthopedic Surgeons, 1995
Poss R: Current concepts review: the role of osteotomy in the treatrnent of osteoarthritis of the hi p. Journal of Bone and Joint Surgery 664:2 44- 15 1, 1984
Takechi H, Nagashima H, ito S: [ntra-articular pressure of the hip joint outside and inside the 1imbus. Jotrrnal of the Japanese Orthopaedic Association 56539-536, 1982
Terayarna K, Takei T, Nakada K: Joint space of the human knee and hip joint under a static load. Engineering in :Medicine 9:67-74, 1980
Weber W. Weber E: Uber die mechanik der menschlichten gehwerkzeuge nebst der beschreibung eines versuches uber das herausfalIen des schenkelkopfes aus der pfame im luftverdennten raurn. Annalen Ph-vsik irnd Chemie 40:l-13, 1837
S.J. Ferguson Queen's University Chapter Two
Background
Labrum Anatomy
Until recently, there has been a paucity of information pertaining to the anatorny oFthe acetabular labrum. Long considered a developmental remnant, in most anatomy textbooks, the acetabular Iabrum is mentioned only briefly, if at all, in the context of overall hip-joint anatomy. For example, even recent editions of Gra-v S Anatomy [5] devote only two lines to the labrum, describing it as, "a fibrocartilaginous lip, which is attached to the bony acetabular margin, deepening the acetabular socket. It is triangular in section and is attached at its base to the acetabular rirn; its apex is its f?ee margin".
For the clinician or engineer seeking specific information about the anatomy of the adult labrum. published studies have been scarce. However, a recent investigation into the anatomy of the labrocapsular complex of the hip joint [85]provides an excellent oventiew of the gros anatomy of the labrum, its morphology and its tissue composition.
The labrum encircles the acetabulum, joining smoothty with the transverse acetabular ligament (TAL) in the inferior portion of the acetabulum. Indeed, it is difficult to visually distinguish the labrum fiom the TAL. The labrum may be separated from the cartilage covered facies hlnata by a thin fissure [85], except in the craniodorsal region and in the region of the acetabular notch, although others have observed no visible border Biomechanics of the Acetabular Labnun
-- --
between the two tissues [19]. in these regions, the articular surfaces of the labrum and
cartilage are smooth and continuous, and the bony structure of acetabular rim steps down
in the same region. In the ventral and dorsal regions, and also in the region of the
transverse ligament, the labrum is approximately 4mrn hi@. Superiorly, at the edge of
the acetabular roof, the iabmrn is thinner and denser.
The Iabrum is triangular in cross section, with approximately a 45 degree angle
behveen the inner (articulating) and outer surfaces. The outer surface of the labrum is
bound to the synovial membrane of the joint, Electron microscopy has been used to study
the morphology of the labrum [83,94]. Shibutani [94]found that its most superficial
layer is a thin membrane characterised by a fine woven mesh of collagen fibrïls.
Immediately below this is a layer, 20 - 100 ym in thickness, with a stratiform structure.
The third, or inner layer, torming the majority of the labrum structure, is composed of thick Type 1 collagen fibre bundles principally arranged parallel to the acetabular rim
(Figure 2-1 ), with some fibre bundles scattered throughout this layer ruming obliquely to the predominant fibre orientation. It is possible that the smaller, oblique fibres within the body of the labrum serve to tie the larger fibre bundles together, reinforcing the structure of the labrum. However, Putz et al [85] found using light microscopy that, in contrast to other fibrous tissue and tibrocartila,oinous structures in the body (cg. the meniscus) there are few loosely packed fibds within the main fibre bundles, and almost no radially oriented fibres. The highly oriented fibres of the labrum appear to forrn a strOng interface with the bony rim of the acetabulum, with individual fibres turning to enter the subchondral bone perpendicular to this interface (Figures 2.2,2.3) [8q.
S.J. Ferguson Queen's University Biomechanlcs of the Acetabular Labnim
Figure 2.1 :Scanning eiecbon rnicroscopic (SEM) image of the labmcross section following keeze-liacturing. The densely packed, circumferentially arranged T-vpe 1 collagen fibres are evident,
Figure 2.2: SEM image of the interface between the labrum and the subchondral bone. Type f collagen fibres enter the subchondral bone perpendicular to the interface. Random orientation of some fibres is an artefact of incomplete freeze hcturing of the fibres ( fiom Owen et al 1831).
S.J. Ferguson Queen's University Biomechanics of the Acetabular Labnun
Figure 2.3: Higher masification image of the interface behveen the labrum and subchondral bone. Collagen fibres are parallel and denseIy packed, forming a strong interface (from Owen et al [83]).
Fibre bundles in the labrum are densely packed and parallel, in contrast to the sparser, mesh-like orientation of the collagen fibres at the cartilage/subchondral bone interface (Figures 2.4,2.5) [83]. The thick Type I collagen fibre bundles of the labrum are also well integrated into the fine Type II collagen network of the adjoining cartilage layer (Figures 2.6 - 2.8) [83]. Tanabe [104] observed fibre bundles 20pm in diameter extending up to 5OOpm from the labrum into the articular cartilage. The same study also demonstrated that the individual collagen tibre bundles of the labrum increase in cross- sectional area through childhood and early adulthood, then decrease in size and density after age thirty. The labrum would appear, therefore, to possess an exceptionalty strong, fibre-reinforced structure quite different from that of the adjoining articular cartilage layer (Figure 2.9) [SI.
SJ. Ferguson Queen7sUniversity Biomechanics of the Acetabular Labrum
Figure 2.4: SEM image of the interface benveen the articular cartilage and subchondral bone of the acetabulum. The fine Type II collagen nehvork of the cartilage matrix exhibits the radial arrangement characteristic of the tissue (fi-om Owen et al [83]).
Figure 2.5: The cartilagehone interface. At higher magnification, the rneshlike arrangement of the collagen fibres is apparent. The collagen fibres at the cartilagehone interface are finer and less densely manged than those at the labrurn/'bone interface (fiom Owen et al [83]) .
S.J. Ferguson Queen's University Biomechanics of the Acetabular Labrum
Figure 2-6: SEM image of the cartilageAabrum interface. Chondrocytes (arrows) are present in both the articular cartilage (AC) and acetabular labrum (AL), although more sparsely in the Iabrum. Thick type 1 collagen fibre bundles can be seen in the labrum (from Owen et al [83]) .
Figure 2.7: Higher magnification of the centre portion of Figure 2.6. A chondrocyte (C) is seen at the interface between the labrurn and cartilage. Thick type 1 cotlagen bundles (1) of the labnim are intertwined with the Fine type II collagen matrix (IL) of the cartilage (fiom Owen et al [83]).
SJ, Ferguson Queen's University Biomechanics of the Acetabular Labrurn
Figure 2.8: Further magnification of the centre portion of Figure 2.7. The fixation method necessary to preserve collagen structure darnages the chondrocyte .
Figure 3.9: Low magnification SEM image of the cross section of the acetabular rim showing the subchondral bone (SB), articular cartilage (AC)and acetabular Iabrum (AL). The iabrum's structure is quite different to that of the adjoining articular cartilage (from Owen et ai [83]).
SJ. Ferguson Queen's University Biomechanics of the AcetabuIar Labrum
- -- -
The composition of the tissue of the labntm remains a topic of debate in the
handfbl of published histological studies of the fetal and adult acetabular labrum. Waiker
et al [12 11 proposed that the labrum is composed of fibrous tissue. The density of
collagen fibre bundles increases as the fetus progresses towards term, but there are no
cartilage-specific cells (Le. chondrocytes) in the labmm. However, Matles [70] claimed
that the acetabular labrum consists of fibrocartilage, not tibrous tissue, at the gross and
rnicroscopic level, and suggested that the failure of others to locate cartilage cells in the
labrum was a result of poor specimen preparation. Putz et al [85] suggested that there are
cartilage cells in the infant labrum, which are related to its role in the development and
ossification of the infant acetabular labrum, but that there is no evidence of cartilage cells
within the adult labrum.
The blood supply for the labrum comes from three locations. A vascular anastomatic ring surrounds the capsular attachent to the acetabulurn, and the blood supply is derived from the superior gluteal vessels, the obturator artery and one ascending branch of the media1 femoral circumtlex artery [30,54,108]. A branch of the nerve to the qiradrutzcs fernoris muscle and the obturator nerve serve for labnun innervation, with many types of mechanoreceptors and free nerve endings in the superficial layers of the labrum [55]. Free newe endings indicate a capacity for pain sensation. while sensory end orsans suggest a proprioceptive capability. Receptor types in the labrurn incl ude Vater-
Pacini corpuscles and Golgi-Mazzoni corpuscles, which are believed to function as pressure sensors, and Ruffmi corpuscles and Krause corpuscles, believed to be temperature sensors. Sirnilar receptors are found in the menisci of the knee [8]. The
S.J. Ferabuson Queen's University Biomechanics of the Acetabular Labrum
innervation of the labrum could explain why labral tears and impingement result in site-
specific hip pain-
With the concentration of these studies on the anatomy and histology of the fetal
and infant labrum, there remain several open questions [41] :eg. how large is the adult
labrum in cross section, hotv is it anchored to the acetabular roof in adults and how does
it derive its nourishment?
Clinical Relevance
There has recently been a marked increase in the number of clinical investigations
into the diagnosis and treatment of labral lesions and their consequences
[4,23,28,3 1,50,5 1,56,7 t ,78,84,102,114,1 t 5,1171. The symptoms of labral lesions include
sharp groin pain, decreased range of motion and a painful instability, or "giving way"
sensation [4,28,3 L,50,56,102]. Diagnosis of labnim lesions is difficult. Magnetic resonance imaging does not consistently identify lesions [56,117], although with
improvements in technique the success of this method is improving 16 1 j. Arthroscopy
rnay be an alternative for both diagnosis and treatrnent of labrum lesions [23,89b, 1021.
The prevailing opinion in the clinical community is that labrum injury is predorninantly a consequence of acetabular rnisalignment, Le. dyspIasia, as the Iabrum must firnction as a secondary stabiliser in the shallow, dysplastic hip, and is therefore subjected to abnormal
levels of tension and shear [28,56,ll4- 1 161. The preferred treatrnent is a realignment osteotomy [28,56,8O,1141 usually in conjunction with a partial excision of the damaged labrum. However, realignment alone without treatment orthe labral defect is not always
SJ. Fer+guson Queen's University Biomechanics of the Ace tabular Labrum
-.-. successful. Gibson and Benson [38] conducted a retrospective study of 12 1 patients treated by removal of the labrum and joint realignment osteotomy. Premanire onset of osteoarthrosis was observed (loss of joint space on radiographs, increased subchondral- bone sclerosis, osteophyte and cyst formation) even in hips that were satisfactorïly reduced. These changes may have been caused by removal of the labrum, despite proper reduction of the joint. Ohzono et ai [80]reported that 50% of patients with a detached labrum proceeded to a poor result following corrective osteotomy, as opposed to 98% with excellent or good results for patients with a normal labrum. Nishina et al [78] ais0 reported an increased chance ofjoint degeneration for patients with a detached labrum.
Damage to the labrum rnay lead to high local stresses and osteophyte formation [28].
Ikeda et al [5 l] postulated that tears in the labrum compromised the load bearing and stability enhancing fùnction of the labrum and, if lefi untreated, would enlarge as a result of repeated stressing. Aithough hip dysplasia is a major contributing factor to labral lesions, there are also reports of labral tears as a result of relatively minor injury in othenvise normal hips [4,3 1 1. LabraI tears alone are often treated by partial labrum resection [89b]. in contrast to dysplastic hips, where labral lesions are mostly in the posterosuperior region, Iabral lesions resulting from injury tend to occur in the antenor and anterosuperior region of the joint. There have been several studies that have shown that labral injury, such as tears or intra-articular impingement, lead to changes consistent with early osteoarthrosis [4,28,45,7 1,791. Cartilage damage has also been reported in areas of labrum tears in othenvise normal hip joints by Rushdfeldt et al [88] in the course of their experimental evaluation of hip joint cartilage layer geometry. Failure of the cartilagehone interface, whereby a flap of cartilage lifts away from the acetabulum, has
SJ. Ferguson Queen's University Biomechanics of the Acetabular Labnun
been observed adjacent to labrum tears [I 9,531, implying increased shear stresses across the cartilagehone interface at the acetabular rim in conjunction with labrum pathology.
Ln a study unrelated to Iabrum injury, Williams et al report on their experience with a new hemiarthroplasty technique [125]. In a hemiarthroplasty, the damaged fernoral head is replaced with a prosthetic device, which articulates in the natural acetabulum. in contrast to cornmon technique, Williams er al preserved the natural labrum in their patients. In their Foliow-up, they reported a decrease in the rate of prosthesis dislocation from 8- 15% (labrurn excised) to just L -3% (labrum preserved).
They observed that the prosthetic head "popped" back into the socket, accompanied by a visible spring of the labrurn tissue as it everted around the fernoral head. They concluded that the labrum significantly improves the stability of the joint.
Etiology of Osteoarthrosis
The etiology of idiopathic - i.e. non genetic, metabolic or traumatic - osteoarthrosis has received much attention in the literature. The Iink between mechanical stress and degenerative changes in articular cartilage has been well documented
[ l6,;7,86,9 1-93-96]. Radin and Paul [86] hypothesized that the high stresses experïenced by the joint are impulsive in nature, and that shock absorption in the joint is accomplished by the bone, rather than by the cartilage. Impulsive loading leads to trabecular fracture, bony rernodelling of the trabeculae and, consequently, stiffer bone. This diminishes the bone's shock attenuating properties, which results in higher cartilage stress and eventually cartilage degradation [87]. However, it is equally Iikely that degradation of
SJ. Ferguson Queen's University Biomechanics of the Acetabular Labnun
the cartilage Ieads to increased peak stresses in the subchondral bone, and is hence a
cause and not an effect of subchondral remodelling, Gardner [35] proposed that a
combination of factors (e.g loading age, metabolic rates) lead to increased cartilage
hydration, disorganisation of the collagen structure, fibrillation and, eventualIy, Ioss of
cartilage.
Articular cartilage is a complex structure, and determination of the actual physical processes that initiate osteoarthrosis and contribute to its progression is not trivial.
Articular cartilage is composed of a composite organic matrix which is swollen by water.
The organic matrix accounts for approximately twenty percent of the total tissue mass amd consists of a meshwork of collagen fibrils, which contains aggregating proteoglycans, glycoproteins, chondrocytes and Iipids [74]. Osmotic potential tends to swell the proteoglycans, with expansion of the proteoglycans limited by tension in the collagen meshwork. Several theories exist for the sequence of events in cartiIage degeneration. but there is general agreement on the underlying rnechanisms. The deveIopment and progression of osteoarthrosis involves disruption or alteration of the cartilage matrix structure and a response of the chondrocytes to these changes [17,35,67-
691. Degradation of the matrix is characterised by fibrillation, cracking, increased water content, stveiling and a decrease in the stiffriess of the tissue. Either as a result of mechanical insult [ 161 or an increase in enzymatic activity, individual col lagen fibrils are split, increasing the perrneability of the tissue as the collagen rnat~xloosens.
Accompanying these changes to the collagenous framework is a decrease in the aggregation of proteoglycans and aggrecan concentration. The initial response of chondrocytes to matrix disruption is an increased level of activity. This repair phase rnay
S .J- Ferguson Queen's University Biomechanics of the Acetabular Labrum
last for years and reverse the course of osteoarthrosis at least temporarily. Enzymatic
degradation by proteoglycanases and collagenases removes both damaged and intact
matrk components. But eventually there is a failure of the chondrocytic response to
restore or maintain tissue levels, either as a result of mechanical damage and death of
chondrocytes no longer protected by an intact matrix, or simply a down regulation of the
chondrocytic response and a loss of balance between the anabolic and catabok processes
maintaining the joint [59]. The end result is a Loss of tissue.
That both mechanical and biochemical factors contribute to the onset and
progression of osteoarthrosis seems clear. However, the mechanism whereby mechanical
loading can lead to matrix disruption has still not been clearly defined. As the
proteoglycan molecules give cartilage its compressive stiffhess, it is clear that increased
rnobility of proteoglycans through collagen matrix failure, and the loss of' large
aggregates directly influence the mechanical integrity of the tissue. To understand these
processes, we must consider the unique composition and consequent mechanics of
cartilaginous tissue.
The Mechan ical Nature of Hydrated Tissues
Hydrated tissues can be likened to a fluid filled open-ceIl sponge. The
collagenous/proteoglycan matrix foms an elastic skeleton, with the voids filled by water.
While lurnped parameter viscoelastic rnodels have been used to descnbe the non-linear, time-dependent behaviour of such tissues under load, a more accurate description of their behaviour is obtained by considering their biphasic (solid/liquid) nature. Biphasic
SJ. Ferguson Queen's University Biomechanics of the Acetabular Labrum
modets have been proposed for articular cartilage [74,100,113], intervertebral discs [95]
and menisci [99]. A load applied to this system will produce a gradua1 creep as the water
is squeezed out of the pores. The rate of settlement depends upon the resistance to flow
through the pores. Such a material is ofien refèrred to as poroelastic (porous, elastic
skeleton) and the material behaviour can be described by Biot's general theory of
consolidation, original1y developed to describe the settlernent of saturated soils [ 131.
Biot's theory States that the total stress within any element of the system is composed of
two parts: one of which is the hydrostatic pressure of the entrained fluid, and the other the
average stress in the elastic skeleton, referred to as the effective stress. The goveming
equations of the elastic-porous medium saturated with a pore fluid, as irnplemented in commercial tlnite element (FE) code, are derived with the following assumptions:
L. the porous medium is fully saturated;
2. the total stresses maintain an elernent of the porous medium in static equilibrium;
3. the effective stress is related to the strains in the skeleton by the elasticity matrix;
4. the solid matrix and pore fluid are incompressible (rate of change in pore volume
equals the rate that fluid is displaced). Note that the tissue, as a whole, is
compressible, due to fluid expression;
5. flow of the pore fluid is soverned by Darcy's law,
The biphasic mode1 proposed by Mow et al [74], based on mixture theory, is identical to Biot's poroeIastic theory if one neglects fluid viscosity. The response of a poroelastic material to load is governed by the elastic modulus, or aggregate modulus, of the drained
SJ. Ferguson Queen's University Biomechanics of the Acetabular Labrum
solid matrix, and the permeability of the matrix. These material properties can be derived from the results of a confined compression test.
The importance of using a poroelastic mode1 for hydrated tissues can be demonstrated in the following figure (Figure 2.10). During the ramp phase (O, A, B) the tissue is compressed at s constant rate by a porous indenter. Fluid is expressed fiom the top surface, and the frictional drag generated by fluid flow requires a large force to cornpress the tissue- During the relaxation phase (C, D, E) the disphcernent of the top surface is maintained.
Fluid Emux and Fluid Solid Flatrix Compaction R~~~~~~~~~~~~~ Equililwium
Figure 3.10: Schematic representation of fluid exudation and redistribution within hydrated tissues during a rate-controlled, compression stress-relaxation experiment. The horizontal bars in the upper figure indicate the strain distribution in the tissue. The lower graph (right) shows the total stress response during the compression phase (0,A, B) and the relaxation phase (B ,C, D, E). (fiom Mow et al [74])
S.J, Ferguson Queen's University Biomecbanics of the Aceîabular Labrum
Since the matnx and fluid are considered incompressible, no tluid is expressed. The
compressive stress at the surface decays as the strain field in the sample equilibrates and
fluid is redistnbuted throughout the tissue. A purely viscoelastic treatment of the tissue
would predict a uniform total strain and stress field throughout the sample, and this is not
the case; the strain field in the solid matrix is highly nonhomogeneous [76].
Recently, Ateshian and Wang [9] showed, using a mathematical mode1 of rolling
contact between cartilage layers, that the fluid pressurisation in contacting hydrated
tissues increases as joint conpence increases, with over 90% of the applied load being
supported by fluid pressure for a significant period of time. In this analysis, a congruent joint was one in which the radii of cwature of the hvo contacting surfaces were
identical, i-e. con*guence at the macroscopic level. This is consistent with the predictions
of Macirowski et al [65]for load support by tluid pressurisation in the hip joint.
Cartilage Failure Mechanisms and Joint Lubrication
An understanding of the Failure mechanism in cartiIage is dificult, as fluid pressurisation leads to tensile stresses in the individual collagen fibrils. While we calculate compressive stresses in the solid matrix, we are really considering a system with proteoglycan molecules being compressed, while tensed collagen fibrils resist their deformation. Although there have been no definitive theories presented, which relate the stresses and fluid pressures predicted by poroelastic theory with the degrdative changes observed in osteoarthrosis, implicit in any discussion of poroelastic tissues is the fact that increasing permeability leads to faster tissue consolidation and higher matrix stresses.
S.J. Ferguson Queen's University Biomechanics of the Acetabular Labrum
High levels of subsurface solid matnx stress may eventually lead to fatigue failure of
cartilage tissue, with intemal failure of the tissue propagating to the material surface
causing cracks and fissures [32,75]. As rhe integity of the collagen matrix determines
the effective "pore" size of the solid matrix, failure of the individual collagen fibrils leads
to higher tissue permeability. This forms a self-driving cycle, whereby the solid matrix
of the tissue will degrade if the rate of damage exceeds the rate of ma& synthesis. This
rnechanism is enhancrd by a decrease in chondrocyte activity, which has been shown to
occur with abnormal stress levels. The collagen fibrils are also responsible for holding
the large proteoglycan molecules in place, and as the collagen fibrils fail, the
proteoglycan molecules are free to move about. It has been speculated that, as cartilage
permeability increases, the preater interstitial fluid flow cm "wash" proteoglycan
rnolecules out of the matrix, further reducins the compressive stiffness of the cartilage
[7,106]. Independent of bulk fluid motion, shearing of cartilage tissue could be a
precursor to matrix disruption.
Not only are the magnitudes of the tissue strains and stresses a factor in
maintenance of normal cartilage function, but also their distribution within the joint. tt
has been postulated that there is some spatial adaptation of cartilage to its mechanical
environment, as the mechanical properties of the cartilage in the hip joint Vary with
location [10,11]. ~hecartilage from highly loaded areas of the joint is generally thicker and stiffer than that From areas that do not normally experience hi& loads. Overloading of cartilage which has been pre-conditioned to lower levels of stress has been suggested as a contributing factor in osteoarthrosis of the patello-femoral joint [93].
S.J. Ferguson Queen's University Biomechanics of the Acetabular Labrum
The customary use of the term "lubrication" may be misleadinj as it is applied to
cartilage. Several rnechanisms have been proposed for cartilage lubrication, and the most
likely seem to be squeeze-film lubrication and weeping lubncation. In squeeze-film
l~brication~as bvo beanng surfaces approach each other, the viscous synovial fluid
cannot be instantaneously squeezed out £tom the intra-articular gap. A hydrostatic
pressure is built up due to viscous resistance to flow out of the gap, which can support
large loads, and distribute the applied load evenly over the entire articulating surface
[75]. Hlavacek [18] used a finite element mode1 to dernonstrate the emcacy of squeeze
film lubrication in synovial joints, and with scanning electron microscopy, Clark et al
[211 found evidence of a fluid film in Loaded diarthrodial joints. Terayama et al [107]
observed fluid layers in the hip joint with thicknesses much greater than those predicted
by theoretical analyses of joint lubrication. In weeping lubrication, the tluid entrained in
the small cartilage surface asperites is expressed between the peaks into the adjoining
vaileys upon contact between articular surfaces. The surface-to-surface sealing is good,
limiting fluid flow into and through the intra-articular gap (Figure 2.1 1). This leads to
pressurisation of the interstitial fluid, which supports the majority of the applied load. In
both cases, friction forces and Wear are reduced by a reduction of the magnitude of
contact solid-on-solid stresses. not by a reduction of the coefficient of friction [62].The
importance of such a mechanism. in which the rnajority of the load applied to a joint is transferred across the cartilage layers through fluid pressure rather than through solid matris stress, has been demonstrated in a variety of theoretical joint contact models
[27.65,90,99,118] and also by direct experimental measurement [98,122]. In the absence of adequate fluid pressurisation, the solid-on-solid stresses and the resulting frictional
S J. Ferguson Queen's University Biomechanics of the Acetabular Labrum
losses inccease. Adhesive, shear-induced Wear due to solid-on-solid cartilage contact has been suggested as a cause of cartilage wear [75], or even temperature-induced damage to the cellular structure of the cartilage itself (163,661.
Figure 1.1 1 : "Weeping" lubrication. (a) flow pattern in a stationary joint under load, Broad arrows indicate flow through the intra-articular space. AL1 arrow lengths represent flow velocity relative to the cartilage matrix. Flow through the intra-articular space is Iimited by sealing (b) isobars for Ioaded joint - applied load is camed predominantly by fluid pressurisation within the tissue (c) flow and isobars during reimbibition of fluid (fiom McCutchen [7la]).
Until it is possible to study the individual collagen fibrii and proteoglycan interactions, we can onIy hypothesize at the response of the solid matrix to mechanical stress. It would seem logical, therefore, to look to high solid matrix stresses as an indication of possible collagen fibd failure, high cartilage solid-on-solid contact stresses
S.J. Ferguson Queen's University Biomechanics of the Acetabular Labrum as an indication of increased frictional heating and adhesive wear, high fluid pressure gradients and fluid velocities as a possible cause of proteoglycan depletion, and high tensile stresses at the IabdcartiIage interface as a cause of tearing and fibrillation.
Biomechanics of Synovial Joints
The topic of hip joint biomechanics has received considerable attention in the past because of its importance in the development, diagnosis and treatment of osteoarthrosis.
Mathematical modelIing, gait analysis, and experiments have been used to characterise load transmission across the hip joint in terms of the total resultant force and also the local mechanical environment. However, the biornechanica1 role of the acetabular labrum has been largely neglected-
Past studies of hip joint articular cartilage have concentrated on joint conpence measurement, contact area measurements and direct pressure measurements- A vanety of radiographic, staininz, casting and direct-measurement studies have attempted to quantify the conpence of the hip joint [l3,18,X,D,40,42, 7 lb, 77,97,120], or the relative fit of the opposing articular surfaces, as the geometry of the cartilage layers greatly influences the local contact stresses within the joint. The general consensus is that the hip joint is slightly incon-ment, with a Iarger femord head than acetabular socket, but the differences in joint femoral and acetabuiar radii are small (clmm). [t is interesting to note that the only non-contact, high-precision measurement of hip joint geometry showed the two joint components to be spherical and con-gruent to the micron level [105]; however, this study was limited to a single specimen. Compared to other diarthrodial
S.J. Ferguson Queen's University Biomechanics of the Acetabular Labnim
joints of the body (the knee, for example), the hip joint is a very con-ment joint, but as will be discussed in Chapter 5, even a relatively small mismatch between the geometry of the two joint surfaces can have a profound idluence on joint function. Radiographie techniques have been used to estimate contact areas, which have then been used to estimate contact stresses, using an appropriate mathematical mode1 [6,14,43,60].
Armstrong er al [6] observed a faster joint consolidation rate for specimens with damaged cartilage, and hypothesised that the increased permeability of the pathological cartilage could lead to fluid transport of proteo~lycans. Day er al [26] measured changes in cartilage deformation behaviour following progressive cartilage sekgment removal to estimate joint contact areas and cartilage stress distribution. The peak stresses calculated exceeded stress levels required to produce fatigue failure in individual cartilage samples.
More recently, intra-articular pressures have been measured directly by transducers mounted below the cartilage surface p,73]. Thin, pressure-sensitive tïlms have been used to measure the static articular pressure distribution in normal and dysplastic hip joints [2,3,29,33,X,72,ll9.l20]. Areas of high pressure (contact stress) often corresponded to observed areas of cartilage degradation. Pressure-sensitive films have also been used to study the changes in joint contact resulting from acetabular fracture
[44,8 1,821, showing that the pressure distribution within the acetabulum is sensitive to changes in the overall geornetry and congmity of the joint. However, Inaba et ol[52] showed that pressure-sensitive films tend to underestimate the actual contact area in the joint, an observation supported by finite-element analysis of contact mechanics [126].
Brown and Shaw [15] mounted piezoresistive transducers in the articular surface to measure the pressure distribution during motion, finding peak pressures in the acetabular
SJ. Ferguson Queen's University Biomechanics of the Acetabular Labnun dome of approximately 9MPa for Ioads of four times bodyweight . Finally, Carlson et al
[2O] instrumented a fernoral endoprosthesis with an array of pressure tramducers and tetemetry and Rushtèldt et ai [89] reported in vitro studies with this device. This device has been implanted in a human subject, allowing the collection of unique in vivo data
[39,49,10 11, and in vitro studies of the post-mortem retrieval have validated the in vivo measurements [22]. Pressure measurements showed high gradients and sipitkant spatial variation in pressure across the articular surface. Also, peak pressures of up to 18
MPa (rising From a chair) were recorded, well above the ultimate strength of the cartilage soIid matrix. This would appear to indicate that the majorïty of the load is supported by pressurised fluid. Pressures in the region of the labrurn were approximately 1.4 MPa during stance phase, versus a peak of 3.1 MPa in the superior region of the acetabulum, so it would seem that a portion of the total joint load is transferred through the labrum.
Although these investigations have contributed meanin-hl data, the experimental methods disrupted Iocal joint con-guity, anatomy and joint lubrication, thus changing the global boundary conditions that influence the local cartilage stresses.
While the hip joint has received considerable attention, experimental investigation of labnrm function is limited to a few published studies. in the Iast century, cadaveric experiments were conducted by Weber and Weber [124] to study the stabiIity of the human hip joint. The hip joint was supported at the margin of a table and the leg was allowed to hang Freety. Afier the rernoval of al1 sol? tissue, including the joint capsule but not the labrum, the joint remained stable. When a hole was drilled through the acetabulum to the articular surface, the fernoral head dislocated. When this hole was sealed with a tïnger, the joint remained stable. After removal of the labrum, the joint was
SJ. Ferguson Queen's University Biomechanics of the AcetabuIar Labnun
no longer stable, and sealing this hole did not prevent dislocation of the femoral head.
They concluded that the labrum provided an "airtight" seal, In a more comprehensive study, Takechi et al [ 1031 rneasured the intra-articular pressure in the hip, using a needle pressure transducer, both inside and outside the labrum. They found that the pressure within the acetabuium was -27 kPa, cornpared to -8 to -13 kPa within the intact joint capsule. The valve effect of the labrum \vas partially compromised when the joint capsule was penetrated. They concluded that the labrum contnbuted to the stability of the hip joint, partly by its "Ieaky" valve effect (sealing of the joint against intra-articular flow of synovial fluid), and partly by the structural resistance to dislocation provided by the tissue itsetf. Distraction of the joint produced a sub-atmospheric pressure in the intra- articular space, which would help to prevent dislocation, but this negative pressure supported only a fiaction of the leg weight. In the absence of other soFt tissues, the remainder of the Ioad must have been supported by the labrum. in these experirnents, the joint load direction \vas opposite to physiological loads and gives only an indication of the stabilising effect of the labrurn seal. A similar sealing phenornenon has been demonstrated for the glenohumeral joint [47,58,123]. However, the capability of the labrum to seal under a compressive, physiologie load has aIso been experimentally proven. Terayama et al [ 1071 dernonstrated that a 0.2 - 0.6 mm thick fluid film remained sealed between the articuIating surfaces of fresh cadaveric hip joints afier applying a
1000 - 1500N compressive load across the joint. In these joints, the labrum remained in tight contact with the opposing femoral cartilage surface (Figure 2.12). Et is important to note that the fluid fiIm observed by Terayarna et al [107] had a thickness orders of magnitude greater than that predicted by, for example, Hlavacek 1481 in analyses OF
S-J. Ferguson Queen's University Biomechanics of the Acetabular Labnun squeeze-film lubrication, raising the possibility oFa lubrication mode unique to the classic squeeze-film mechanism.
Figure 2.12: Magnified picture of the peripheral portion of the hip joint under a load of 1275N. The Iabrum (L) appeared to seal a fluid Iayer between the cartilage surfaces. The articular cartilage under the labrum was markedly depressed (from Terayama er al [107])
In their study of the structure and Function of the transverse acetabular ligament, Lohe et al [64] concluded that the labnim did not contribute significantly to the reinforcement of the acetabular rim, due to the low strain levels measured during ex vivo joint loadiny.
However, the results of this same study are cited to support the hypothesis that the labrum is indeed a structural element of the acetabulum [as]. Recently, Konrath er ul [57] studied the role of the acetabular labrum and transverse ligament in load transmission over the hip. Using pressure sensitive film, they measured the changes in local contact pressure within the hip joint during simulated single legged stance following labrum
S.J. Ferguson Queen's University Biomechanics of the Acetabular Labnrm
excision, The only significant difference detected was a decrease in maximum pressure in the posterior region of the acetabulum. However, it should be noted that the accuracy of pressure sensitive film for determining joint contact pressures is an open question in the biomechanics community [52,126]. Tt is possib!e that the introduction of this film into the intra-articular space alters the joint sealing properties. K~nrathet al [57] acknowledged that the labnrm's contribution to stabilisation, proprioception, shock absorption or lubrication of the hip joint is still unknown,
Mathematical and finite element models have been used to predict, rather than directly measure, joint contact pressures as well as providing additional information which cannot be easily obtained fiorn in vitro experiments. Severai whole joint models have been developed. Knee joint models have been used tc study the load transmitting role of the tibrocartilaginous menisci, with boundary conditions derived from load- displacement rneasurements of intact joints. Early models treated the soft tissues as
Iinearly elastic, isotropic materials [46]. Tissakht et al [Il21 refmed their knee model by treating the menisci as composite structures consisting of a membrane covered matrix.
The addition of transversely isotropic material properties and three-dimensionaI geometry further improved the model [log- 1 1 11. They demonstrated that meniscal injury fiom pure compression is not likely, but that the addition of a rotation results in tende stresses higher than the ultimate tensile strength of collagen fibrils in the centre-posterior regions.
This corresponds to the location with the highest incidence of bucket-handle rneniscal tears. Spilker et al 1991 used a poroelastic formulation to model the rnenisci. Their model showed that the majority of the load is carried by fluid pressure. They also observed that their model predicted large tensile stresses in the periphery of the menisci.
S.J. Ferguson Queen's University Similar poroelastic models have been developed to study the behaviour of intervertebnl discs [95].
Genda et al [36] used a linear elastic three-dimensional FE mode1 of the hip to investigate contact pressures in normal and dysplastic hips. AP radiographs were projected into three dimensions, assuming a spherical acetabulum, and the cartilage was modelled as a series of individual elastic springs. The study showed that contact pressures increased greatly with decreased lateral and anterior coverage. Dalstra er al
[24] developed a sophisticated three-dimensional model of the pelvis to study load transfer across the hip joint. Their model was validated against experimentally measured strains in the cortical shell of the pelvis. However, the model did not include the cartilage layers of the Femoral head and acetabulum and so offered no information about the stress state within these sofi tissues. Macirowski er al [65] developed a three-dimensional poroelastic model of the hip, the only rnodel to treat the cartilage layers of the hip as a hydrated tissue. Experimental data fi-om consoIidation experiments - pressures measured with an instrumented endoprosthesis - were used as boundary conditions for the model, and the corresponding fluid pressures, fluid velocities and soLd ma& messes were calculated. The model demonstrated, once again, the importance of fluid pressurisation within the cartilage layers in load transfer across the hip. They hypothesised. from the large spatial variation in pressure rneasurernents and slow consolidation nte, that the cornpliant articular surfaces seal almost immediately upon contact, preventing fluid escape into the articular gap, and it is this surface to surface sealing, which is critical for low wear in the hip joint. However, the experiment and finite element madel did not consider the influence of the labrum on the overall joint consolidation. While the hip
S.J. Ferguson Queen's University Biomechanics of the Acetabular Labrum
joint itself has received considerable attention, there are no published finite element analyses of labrum hnction.
References
Adams D, Kempson GE, Swanson SA: Direct measurement of local pressure in the cadaveric human hip joint, :Med-BioLEng-Compzlr. 16: 1 13-1 15, 1978
Moke NY, Byers PD, Hutton WC: Contact pressures in the hurnan hip joint. Jozrrnaf of Bone and Joint Swgery 69-A:536-54 1, 1987
Ahmed AM, Burke DL: ln-vitro rneasurement of static pressure distribution in synovial joints - Part 1: Tibia1 surface of the knee. Jorrrnal of Biomechanical Engineering 1O5:Z 16- 225, 1983
Altenberg AR: Acetabular labrum tears: a cause of hip pain and degenerative arthritis: Sozrther-il Medical Jorrrnal70 :1 74-1 75, 1977
Anonymous: Structure of synovial joints. In: Grqv's Anatomy-, 36th, pp PL Williams and P Warwick. Edinburgh, Churchill Livingston, 1980
Armstrong CG, Bahrani AS. Gardner MA: In-Vitro measurement of articular cartilage deformations in the intact human hip joint under load. Jozrrnal of Bone and Joint Szrrget;~ 6 l-A:7U-75j7 1979
Armstrong CG, Mow VC: Variations in the intrinsic mechanical properties of human articular cartilage with age, degeneration and water content, Jorrrnal of Bone and Joint Szrrgery 64-A:8 8-94, 1982
Assimakopouluos AP, Katonis PG, Agapitos MV, Exarchou EI: The innervation of the human rneniscus. Clinical Orthopaedics and Refared Reseurch 275:23 3-23 6, 1 992
Ateshian GA, Wang H: A theoretical solution for the frictionIess rolling contact of cylindrical biphasic articular cartiiage layers. Joztrnal of Biomechanics 28: 134 1-1 355,
Athanasiou KA, Agarwal A, Dzida FJ: Comparative study of the intrinsic mechanical properties of the hurnan acetabular and femoral head cartilage. Jozirnnl oforthopaedic Resea~ch1 2:MO-WI. 1994
[LI] Athanasiou KA, Agarwal A, Muffoletto A, Dzida FJ, Constantinides G, Clem M: Biomechanical properties of hip cartilage in experirnental animal modeIs. Clinical Orrlzopnedics and Relded Research 3 16:254-266, 1995
[ 121 Biot MA: General theory of three-dimensional consolidation. J.Appf.Phys. 12: 155- 164, 1941
S.J. Ferguson Queen's Universis Biomechanics of the Acetabular Labrurn
Bliisius K, Weiss M, Beks P, Neusel E, Niethard FU: Macromorphological comparison of fiozen sections concerning the problem ofconpency oFthe hip joint. Acta Orthopaedica Trarima Sur-gety 1 I4:83-86, 1995
Brinckmann P, Frohin W, Hierholzer E: Stress on the articular surface of the hip joint in healthy adulrs and persons with idiopathic osteoarthritis of the hip joint. Jorirnal of Biomechanics 14: 1 49-1 56,198 1
Brown TD, Shaw DT: in vitro contact stress distributions in the natural human hip. Jorrrnal of Biomechtics 16573-384, 1983
Buchval ter J A: Osteoarthritis and articular cartilage use, disuse and abuse: experimental studies. The Jorrrnal of Rheumatology 22: 13-15, 1995
Buchvalter .JA, Mankin EU: Articula cartilage. Part 11: degeneration and osteoarthrosis- Repair, regeneration and transplantation, Jorirnal ofBone andJoinr Surgey 79-A:6 12-632, 1997
Bullough PG, Goodfellow J, Greenwald AS, O'Connor J: Inconopent surfaces in the human hip joint. Nature 2 17: 1290, 1968
Byers PD, Contempomi CA, Farkas TA: A post mortern study of the hip joint including the prevatence of the katures of the right side. Amals of rhe Rherirnatic Diseases 29: 15, 1970
Carlson CE. Mann RW, Hams WH: A radio telemetry device for monitoring cartilage surface pressures in the hurnan hip. iEEE Trans.BiornedEng-2 1257-264, 1973
Clark JM, Norman AG, Kaab MJ, Notzli HP: The surface contour of articular cartilage in an intact. loaded joint. LAnat. 195:45-56, 1999
Clayton iMC, Carlson CE, Hodge WA, Mann RW: Postmortem cartilage changes due to in vivo hip pressures. Proceedings of rhe 38th Annzral ,Meeting of the Orthopaedic Resenrch Society, Washington, DE., pp- 6 13, 1992
Conn KS, Villar RN: [The tom acetabular Iabrum - a hip arthroscopist's view]. Orthopüde 27:699-703, 1998
DaIstra MlHuiskes R, van Eming L: Development and validation of a three-dimensional finite element mode1 of the pelvic bone. Joirrnal of Biornechanical Engineering 1 l7:?7Z 378, 1995
Davis PH, Frymoyer JW: A femoral head measuring device. Journal of Bone nndJoint Srirgery 5 1-A: 1663-1664, 1969
Day WH, Swanson SA, Freeman MA: Contact pressures in the loaded human cadaver hip. .formal of Bone and Joint Surgey 57-A:302-3 1 3, 1975
Donzelli PS, Ecqht,r,inF, Putz R, Spilker RL: Physiological joint incongruity significantly affects the load partitioning between the solid and fluid phases of articuIar cartilage. Proceedings of the 13rd Annzral Meeting of the Orthopaedic Research Sociew, San Francisco, California, pp. 82, 1997
S.J. Ferguson Queen's Universi5 Biomechanics of the Acetabular Labrurn
Dorrell JH, Catterall A: The rom acetabular labrum. Jozirnal of Bone andJoint Sttrgery [Br] 68-B:400-403, 1986
Eckstein F, von Eisenhart-Rothe R, Landgraf J, Adam C, Lohe F, Müller-Gerbl M, Putz R: Quantitative analysis of incongruity, contact areas and cartilage thickness in the human hip joint. Acta Anatomica 1 58: 193-204, 1997
Fischer LP, Noyer D, Gonon GP, Carret JP, Morin A, Clermont A: [Artenal vascularization of the os coxae]. BrrlLAssocAnat. 6 1343-356, 1977
Fitzgerald EW: Acetab ular Iabrum tears. Chtical Orthopaedics and Related Research 3 1 1 :60-68, 1995
Freeman MAR: Is collagen fatigue failure a cause of osteoarthrosis and prosthetic component migration? A hdvpothesis.Jorirnal of Orthopaedic Reseçrrch 173-8, 1999
Frisina W, Lehneis HR: Pressure mapping: a preliminary report. Journal of Biornechanics 3536, 1970
Fukubayashi T, Kurosawa H: The contact area and pressure distribution pattern of the knee. A study of normal and osteoarthrotic knee joints. Acta Orthopaedica Scandinavica 5 1 :87 1-879, 1980
Gardner DL: The nature and causes of osteoarthrosis. Brit. Med-J. 286:4 18-424, 1983
Genda E, Konishi N,Hasegawa Y, Mima T: A computer simulation study of nomal and abnormal hip joint contact pressure. Arcll.Orthop. Trarirna Srirg- 1 14:202-206, 1995
Ghosh P, Read R, Armstrong S, Hoibert C, Sun D: High contact stress indiced by rneniscectorny results in cartilage and s-vnovial fluid changes consistent with earIy osetoarthritis - a 2 year study in sheep. Proceedings ofrhe Combineci Mkering of the Orthopaedic Research Society, San Diego, Cal ifomia, pp. 57, 1 995
Gibson PH, Benson MK: Congenital dislocation of the hip. Review at maturity of 147 hips treated by excision of the limbus and derotation osteotomy. Jorrrnal of Bone andloint Srirge~[Br] 64-B: 16%1 75, 1 982
Givens-Heiss DL, Krebs DE, Riley PO, Strickland EM, Fares M, Hodge WA, Mann RW: In vivo acetabular contact pressures during rehabiiitation, part II: postacute phase. Ph-vsical Therapy 72700-705, 1992
Goodfellow J. Mitsou A: Joint surface incongruity and its maintenance. Jozirnal ofBone and Joint Srirgery [Br] 59-BA4645 1, 1977
Graf R: [The acetabular labrum and labrum lesions]. Orthopade 27:669, 1998
Greenwald AS, Haynes DW: Weight bearing areas in the human hip joint. Jorir-nal of Bone and Joint Stirgey 54-A:157-163, 1972
SJ. Fermouson Queen's University Biomechanics of the Acetabular Labrum
HadIey NA, Brown TD, Weinstein SL: The effects of contact pressure elevations and aseptic necrosis on the long-term outcome ofcongenital hip dislocation. Jorrrnal of Orrhopaedic Research 8 :5O4S 13, 1990
Hak DJ, Harnel AJ, Bay BK, Sharkey NA, Olson SA: Consequences of transverse acetabular fracture malreduction on load transmission across the hip joint. J-Orthop.Traztnta l2:9O- 100, 1998
Hams WH, Bourne RB, Oh 1: Intra-articula acetabular labrum: a possible etiological factor in certain cases of osteoarthritis of the hip. Jorrrnal of Bone and Joint Sz~rgery6 1- A:5 10-5 14, 1979
He@ MS, Zoghl -M: The menisci: a finite element model. Proceedings ofthe 34th Annzral Meeting of the Orthopaedic Reseurch Society, Atlanta, Georgia, pp. 1 49, 1988
Helmig P, Sojbjerg JO, Sneppen O, Loehr JF, Ostgaard SE, Suder P: Glenohumeral movement: patterns aAer puncture of the joint capsule: an expenmental study. Jorrrnal of Shozrlder and EZborv Strrgery 2209-2 15, 1993
Hlavacek M: The role of synovial fluid filtration by cartilage in lubrication of synovial joints - IV. squeeze-film lubrication: the central film thickness for normal and inf-larnmatorysynovial fluids for axial symmetry under high loading conditions. Jotirnal of Biomechanics 28: 1 199- 1205. 1995
Hodge WA, Carlson KL, Fijan RS, Mann RW: Contact pressures fiom an instrumented hip prosthesis- Jownal of Bone andJoint Surgery 7 1 -A: 1378- 1 3 86, 1989
Hofmann S, Tschauner C, Urban M, Eder T, Czerny C: [Clinical and irna+$ng diagnosis of lesions of the labrum acetabulare]. OrrhopGde 37:68 1-689, 1998
ikeda T. Awaya G. Suzuki S. Okada Y. Tada H: Tom acetabular labrum in young patients. Arthroscopic diaposis and treatment. Jazrrd of Bone andJoinr Szrrgery [Br] 70-B: 13- 16, 1988
Inaba H, Ani M: A method for measuring contact pressures instantaneously in articular joints- Jolrrnal of Bionrecl~nnics22: 1 293- 1 296, 1989
Ito, K.: Personal Communication, 1999
Katthagen BD, Spies H, Bachmann G: [Arterial vascularization of the bony acetabulum]. Zeitschrifi frrr Orthopadie zrnd ihre Grecgebiere 13 3 :7- 1 3, 199 5
Kim YT, Azuma H: The nerve endings of the acetabular labrum. Clinical Orthopaedics andRelaredResecrrch 330:176-181, 1995
Klaue K. Durnin CW, Ganz R: The acetabular rhn syndrome. A clinical presentation of dysplasia of the hip. Jortrnal of Bone andJoint Swgery [Br] 73-B:423-429, 1991
Konrath GA. Harnel Ai, Olson SA, Bay B, Sharkey NA: The role of the acetabular labrum and the transverse acetabular ligament in Ioad transmission in the hip. Journal of Bone and Joint Srrrgery 80-A: 178 1-1 787, 1998
SJ-Ferguson Queen's University Biomechanics of the Acetabular Labrum
[58] Kumar VP, BaIasubramaniam P: The role of atmospheric pressure in stabilising the shoulder. Jozmal of Bone and Joint Srrrgery (Brl67-B:7 19-72 1, 1985
[59] Lanzer WL, Komenda G: Changes in articular cartilage afier men iscectomy, Chical Orthopaedics and Related Research 35241 48, 1 990
[60] Legal H: introduction to the biomechanics ofthe hip. ln: Congenital Dysplasia and Dislocation of the Hip., D Tonnis. Berlin, Springer-Verlag, 1987
[6 11 Leunig M, Werlen S, Ungersbock A, Ito K, Ganz R: Evaluation of the acetabular labrum by MR arthro+;nphy. Jorrrnal of Bone and Joint Szrrgery [Bd 79230-234, 1997
[62] Lewis PR, McCutchen CW: Mechanisrn of animal joints. Nature 1 85?: 1284- 1385, 1959
[63] Lindquist S: The heat-shock response. Annual Reviav of Biochemisrty 55: 1 15 1-1 19 1, 1986
[64] Lohe F, Eckstein F, Sauer T, Putz R: Structure, strain and fwiction of the transverse acetabular ligament. Acta Anarornica 157:3 1 5-323, 1996
[65] Macirowski T, Tepic S, Mann RW: Cartilage stresses in the human hip joint. Journal of Biomechanical Engineering 1 1 6:1 0- 18, 1994
[66] Madreperla SA, Louwerenburg B. Mann RW, Towle CA: Induction of heat shock protein synihesis in chondrocytes at physiological temperatures. Journal of Ot-thopaedic Reseorch 3:30-35, 1985
[67] Mankin HJ: The reaction of articular cartilage to injury and osteoarthritis (ktof hvo parts). New Engiond Jozcrnal of hfedicine 29 1 :1 285- 1 292, 1974a
[68] Mankin Hi: The reaction of articular cartilage to injury and osteoarthritis (second of two parts ). :%w Englund Jotlrnal of'kfedicine 29 1: 1335- 1340, 1974b
[69] Mankin HJ: Curent concepts review. The response of articuIar cartilage to mechanical injury. Jozirnal of Bone and Joint Szrrgery 64-A:460-466, 1 982
[70] Matles AL: A rnicroscopic study of the newbom fibrocartilagenous acetabular fabrum. Clinical Orthopaedics and Related Resenrch 54: 1 97-206, 1 967
[7 11 McCarthy JC, Busconi B: The role of hip arthroscopy in the diaposis and treatrnent of hip diseass, Canadian Journal of S1iuget-y 3 8 Suppl 1 :S 13-1 7, 1995
[7 la] McCutchen, CW: An approximate equation for weeping lubrication, solved with an electrical analogue, Annals of Rherrrncrtic Disease 34:85-90, 1975
[7 1 b] Menschik, F: The hip joint as a conchoid shape. Jorrrnul of Biomechanics 30:97 1-973, 1997.
[72] Michaeli DA, Murphy SB, Hipp JA: Cornparison of predicted and measured contact pressures in normaI and dysplastic hips. Medical Engineering and Physics 19: 180-1 86, 1997
S.J. Ferguson Queen's University Biomechanics of' the Acetabufar Labrum
Minahi J. Solomon L. Kaufman B, Duggan TO: An experirnental method for investigating load distribution in the cadaveric human hip. hrrmal of Banc andJoint Sir-gery [Berq 63- B:610-613, 198 1
Mow VC, Kuei SC, Lai WM, Armstrong CG: Biphasic creep and stress relaxation of articular cartilage in compression: theory and experiments. Jorrrrial ofdiot~techanical Engineering 10273-84, 1980
Mow VC, Soslowsky LJ: Lubrication and wear ofjoints. in: Basic orthopaedic biontechanics, 245-292. Ed by VC Mow and WC Hayes. New York, Raven Press, 1991
Mow VC, Zhu W, Ratclittè A: Structure and fùnction of articular cartilage and meniscus. Ln: Basic Orthopaedic Biomechanis, 1. pp 143-198. Ed by VC Mow and WC Hayes. New York, Raven Press, 199 1
Neusel E, Bierther B, Fuss B, BlZsius K: Hip joint conPence: experimental studies. Archives of Orthopaedic Trazrrna Swgev 1 15332 1-324, 1996
Nishina T, Saito S, Ohzono K, Shimizu N, Hosoya T, Ono K: Chiari pelvic osteotomy For osteoarthxitis. Jorrrnal of Bone clndJoint Srtrgeery [Br] 72-B:765-769, IWO
Oh& H: [An arthroscopie study ofcoxarthrosis]. Nippon Seikeigeka Gakkoi Zasshi 68:125-138, 1994
Ohzono K. Sakai T. Haraguchi K. Lee K, Matsui M, Sugano N, Nishii T, Ochi T: vhe Osaka experience. Dome osteotomy with and without labral resection]. Orrhopude 27:759- 764, 1998
Olson SA, Bay BK, Chapman MW, Sharkey NA: Biomechanical consequences of fracture and repair of the posterior wall of the acetabulum. Joitrnol ofBone andJoint Swgen; 77- A:lI84-1192. 1995
Olson SA, Bay BK, Pollak AN, Sharkley NA. Lee T: The effect ofvariable size posterior wall acetabular iiactures on contact characteristics of the hip joint. J-Orthop.Trauma 10:395-402, 1996
Owen GWi, KZib MJ, Ito K: Scanning electron microscopy of coIlagen netsvork morpholog at the cartilage, labrum and bone interfaces in the acetabulum. Scaming :WCI-oscopyInternotional Suppl 1 2: 1999
Paterson 1: The tom acetabular labrum, A block to reduction of a dislocated hip. Joztt-nal of Bone and Joint Strrgery [Br] 39-B:306-309, 1957
Putz R Schrank C: [Anatomy of the labrocapsular cornplex of the hip joint]. Orrhopade 37:675-680, 1998
Radin EL, Paul IL: Response ofjoints to impact loading. 1. in vitro wear. Arthritis & Rhelrrnatisrn 14356-362. 197 1
Radin EL, Rose RM: Role of subchondral bone in the initiation and progression of cartilage darnage. Clhical Orthopoedics and Related Research 2 1 3 :3WO, 1 9 86
SJ. Ferguson Queen's University Biomechanics of the Acetabular Labrum
[88] Rushfeldt PD. Mann RW, Hm-s WH: improved techniques for measuring in vitro the geometry and pressure distribution in the human acetabulum - 1- ultrasonic measurement of acetabular surfaces, sphericity and cartilage thickness. Jorirnal of Biomechanics 13:253- 260,1981a
[89] Rushfeldt PD, Mann RW, Hams WH: Improved techniques for measuring in vitro the geornetry and pressure distribution in the human acetabulum - 11- instnunented endoprosthesis measurement of articular surface pressure distribution. Jozrrnal of Biomechanics 14:3 15-323, 198 1b
[89b] Santori N, Villar RN: Acetabular labrum tears: results ofarthroscopic partial limbectomy.
Schreppers GJ, Sauren AA, Huson A: A model of force transmission in the tibio-femoral contact incorporating fluid and mixtures. Proceedings of the Institution of Mechanical Engineers. Journal of Engineering in bfedicine [KI 205:233-24 1. 199 1
Seedhom BB: Transmission of the load in the knee joint with special reference to the role of the menisci, Part 1: anatomy, analysis and apparatus. Engineering in -Medicine 8207- 217. 1979a
Seedhom BB: Transmission of the load in the knee joint with special reference to the role of the menisci, Part II: experimental results, discussion and conclusions. Engineeriug in Medicine 8:220-328, 1979b
Seedhom BB, Takeda T, Tsubuku M, Wright V: Mechanical factors and pateHofemora1 osteoarthrosis. Annafs of rhe Rl~errmaricDiseases 38507-3 16, 1979
Shibutani N: Three-dimensional architecture of the acetabular labrum - a scanning electron microscopie study. Joro-na?of the Japanese Ortlzopaedic Association 6232 1-29? 1988
Simon BR, Wu JSS, Carlton MW, Evans JH, Kazarian LE: Structural models for human spinal motion sebments based on a poroelastic view of the intervertebrai disk. Jorirnal of Biomechanical Engineering 1O7:32?-33 5, 1985
Simon SR, Radin EL, Paul IL, Rose RM: The response of joints to impact loading, II. In vivo behaviour of subchondral bone. Jozrr-nal ofBiomechmics 5267-272, 1972
Simon WH, Friedenberg S, Richardson S: Joint congruence: a correlation ofjoint congruence and thickness of articula cartilage in dogs. Jozrrnal of Bone andJoint Srrrgery 55-A:1614-1620, 1973
Soltz MA, Ateshian GA: Experimental verification and theoretical prediction of cartilage interstitial fluid pressurization at an impermeable contact interface in confined compression- Jorirrzal of Biomechanics 3 1 :927-934, i 998
Spilker RL, Donzelli PS, Mow VC: A transversely isotropic biphasic finite element model of the meniscus. Jozrrnnl of Biornechanics 25:1027-1045, 1992
[100] Spilker RL, Suh JK, Mow VC: A finite element formulation of the nonlinear biphasic mode1 for articular cartilage and hydrated sofi tissues including straindependent
S-J. Ferguson Queen's University Biomechanics of the Acetabular Labrum
permeability. ln: Comprrtational Methods in Bioengineering., 81-92. Ed by RL Spilker and BR Simon- New York, Amencan Society of Mechanical Engineers, 1988
[IO11 Strickland EM, Fares M, Krebs DE, Riley PO, Givens-Heiss DL, Hodge WA. Mann RW: In vivo acetabular contact pressures during rehabilitation, part 1: acute phase. Physical Therap+v72:69 1-699, 1992
[IO21 Suzuki S, Awaya G, Okada Y, Maekawa M, Ikeda T, Tada H: Arthroscopic diagnosis of ruptured acetabular labrum. Acta Orrhopaedicn Scandinavica 57513-5 15, 1986
[103] Takechi H, Nagashima H, Ito S: Intra-articula pressure of the hip joint outside and inside the limbus. Jorirnal of the Jnpmese Orthopaedic Association 56529-536, 1982
[104] Tanabe H: Aging process of the acetabular Iabrum - an eIectronic microscopie study, Jorrrnal of the Jupanese Ortl-opaedicAssaciution 65:1 8-25, 199 1
[IO51 Tepic S: Congniency of the himan hipjoint, MSc, Thesis, Massachusetts Institute of Techn~lo~q.l98Oa
[ 1061 Tepic S : CS;nantics of and entropy prodzrcrion in the cartilage layers of the synovialjoint-, Sc-D. Thesis. Massachusetts Institute ofTechnology. 19S2b
[107] Terayama K, Takei T, Nakada K: Joint space of the human knee and hip joirit under a static foad. Engineering in :Medicine 9:67-74. 1980
[ 1081 Thomas G: [Arterial vascuIar supply of the upper acetabular region in adults]. At-clz.Orthop. Unfallchir. 58300-305, 1965
[ 1091 Tissakht M. Ahmed AM: Effect of axial rotation on knee menscal stress: a finite element study. In: Proceedings ofthe 36th annual meeting ofthe Ot-rhopaedic Research Sociep. Febrtiary 5-8. /990, New Orleans, Louisians, 1990
[1 101 Tissakht M, Ahmed AM, Misra AK: Meniscal response to 3D comples loadings using a finite element model of the intact tibiotèmoral joint. Proceedings of the 38th Annriaf Meeting of the Orthopaedic Resenrch Society, Washington, D.C., pp. 628, 1992
[Ill] Tissakht M. Ahmed AM, Tanzer M, Misra AK: The thickness of the articular cartilage affects the load-bearing characteristics of the menisci. Proceedings of the 40th Annziaf Meeting of the Orthopaedic Research Society, New Orleans, Louisiana, pp. 41 4, 1994
[ 1 121 Tissakht M. Marchand F, Ahmed AM: Non-linear finite element analysis of the knee menisci: a composite fiber-reinforced model. Proceedings of the 3 7th Annzral ;Meeting of the OrthopaecficResearch Society, Anaheim, California, pp, 294, 199 1
[ 1 131 Torzilli PA. Mow VC: On the fündamental fluid transport mechanisms through normal and pathological articular cartilage during function - II. the analysis, solution and conclusions. Jorrrnal of Biornechanics 9:587-606, 1976
[ 1 141 Tonnis D, Kalchschmidt K, Heinecke A: [Acetabular rotation by triple pelvic osteotomy according to Tonnis]. Orthop&fe37:733-742, 1998
S.J. Ferguson Queen's University Biomechanics of the Acetabular Labrum
[Il51 Tschauner C, Hofmann S: [Lesions of the acetabular labrum in residual dysplasia of the hip joint, Biomechanical considerations regarding pathogenesis and treatment]. Orthopade 27:725-732, 1998
[Il61 Tschauner C, Hofmann S, Graf R, Engel A: [Labrum lesion and residuai dysplasia in the hip]. Orrhopüde 27:772-778, 1998
[Il71 Urban M, Hofmann S, Tschauner C, Czerny C, Neuhold A, ffimer J: [hm arthrography in lesions of the [abmm acetabulare, Technique and role]. Orrhopiide 27:69 1-698, 1998
[II81 van der Voet AF, Shrive NG, Schachar NS: Numerical modelling of articular cartilage in synovial joints - poroeIasticity and boundary conditions. In: Cornpzrter :tl;7tlrods in Biomechanics and Biomedical Engineering, 200-209. Ed by J Middleton, GN Pande, and KR Williams. Swansea, IBJ Publishers, 1993
[119] von Eisenhart-Rothe R, Eckstein F, Landgaf J, L6he F, Müller-Gerbl M, Putz R: Verteilung der anatomischen gelenkspaltweite und flachenpressung irn menschlichen hüftgelenk - eine quantitative analyse. Osteologie 5:55-63, 1996
[120] von Eisenhart-Rothe R, Eckstein F, Müller-Gerbl M, Landgraf J, Rock C, Putz R: Direct cornparison of contact areas. contact stresses and subchondral mineralization in human hip joint specimens. Acta Ernbpolog- l95:279-288, 1997
[ 12 11 Walker JM: Histological study of the tètal development of the human acetabuhm and labrum: sigificance in congenital hip disease. Yale J.Biol.Med 54255-263. 198 1
[122] Wang LH, Soltz MA, Ateshian GA: InterstitiaI fluid pressurizarion regdates the fiictional response of cartilage. Proceedings of the 13r.d Annzraf Meeting of the Orthopaedc Research Society, San Francisco, California, pp. 83, 1997
[123] Warner JJP, Deng X, Warren RF. Torzilli PA, O'Brien SJ: Superohferior translation in the intact and vented glenohumeral joint. Jorrrnal of Sho~rlderand Elbow Srrrgey 299-1 05, 1993
[124] Weber W, Weber E: Uber die mechanik der menschlichten gehwerkzeuge nebst der bsschreibung sines versuches uber das herausfallen des schenkeIkopfes aus der pfanne im luflverdennten raum. Annalen Physik und Chernie 40: 1-1 3, 1837
[ t 251 Williams CRP,Kemohan JG, Sherry PG: A more stable posterior approach for hemiarthroplasty of the hip. Injtqv 38279-28 1, 1997
[126] Wu JZ, Herzog W. Epstein M: Effects of inseriing a pressensor film into articular joints on the actual contact mechanics. Jorrrnal of Biomechanical Engineering 130:655-659, 1998
S.J. Ferguson Queen's University Chapter Three
Three-dimensional Computational Reconstruction of Mixed Anatomical Tissues following Histological Preparation
S.J. ~er~uson'*',J.T. £3ryant3and K. 1toI2
'AO ASIF Research Institute, Davos Platz, Switzerland
'~e~artmentof Orthopaedic Surgery, University of Berne, Inselspital. Berne, Switzerland
3 Department of Mechanical Engineering, Queen's University, Kingston, Canada
Published in: Medical Engineering and Physics (1 99 P), 2 1: 1 il - 11 7 Biomechanics of the Acetabular Labrum
Abstract
The creation of geometrically accurate cornputer models of anatomical structures
with complex shape and mixed tissue types can be difficult. A method for shape
reconstruction based on digital images of polyrner embedded, serially sectioned
specimens is presented. The distortion of bone and sofl tissue specimens during al1
stages of histological preparation was measured. Serial sections of one specimen were
stained with ccmmon histo10,aical stains to enhance the contrast between different tissue
types. High-resolution digital images O f these sections were then processed into a three-
dimensional solid mode1 usuig commercial sofbvare. Preparations containing bone and
cartilaginous tissues were dimensionatly stabIe folIowing fixation, dehydration and
embedding (shrinkage / expansion less than 2%). Staining was necessary to identify
anatomka1 features that othenvise could not be di fferentiated from their surroundings.
Although time consuming, this method provides cross section images of a higher resolution than those obtained fiom CT or MN scanning, and with better sofi tissue
visualisation.
Keywords: 3D reconstruction, anatomy, histology
S.J. Feraouson Queen's University Biornechanics of the Acetabular Labrurn
Introduction
In the field of biomechanics, it may be necessary to create geometrically accurate cornputer models of anatomical objects with a higher resolution and better soft-tissue differentiation than that available with current techniques. Accurate anatomical data of objects with complex shape or composition forms the basis of computational models used in kinematic analysis and finite element analysis (FEA). Whole-joint models, which incorporate both bone and sofl tissues, could be used to study the interaction of anatomical structures, for example the effèct of the glenoid labrum on the stability and range of motion of the shoulder joint, Three-dimensional reconstruction from serial cross-section images is the most common technique for creating such a model, For most
FEA studies based on bone anatomy, computed tomo-mphy (CT) scan datasets are used to define the model geometry, Three dimensional models based on CT scan data have been used extensively to study the femur, pelvis and spine [8,1 1,14,18,19]. High resolution images of trabecular bone structure have been obtained using microscopic CT
[12], but at present this method is limited to small specimens. The primary advantages of the CT method are: minimal distortion of the bone cross section on the CT image, automated image processing for shape extraction, in vivo cross-sectional data, and the possibility of obtaining additional information about bone stifhess. However, it is less desirable to use CT data as the basis for rnodels, which include structures other than bone, due to the poor resolution of soft tissues. Magnetic resonance imaging (MM) can be used to obtain cross-sectional images with better differentiation of soft tissues [Il, but the resolution of MR images is Iimited, typically 256 x 256 pixels. In many cases,
S.J. Ferguson Queen's University Biomechanics of the Acetabular Labntm
depending on the sophistication of the irnaging sequence used. discrete anatomical
structures, such as the acetabular labrum and adjoining articular cartilage, cannot be
delineated due to the ambiguous relationship between signal intensity and tissue
composition [9].
CT MRI Frozen Section
Figure 3.1 :Cornparison of sectioning techniques for large specirnens. The images shown are magnifications of a portion of whole body sections fiom each method and provide a cornparison of the relative resolution. CT-scans provide hi& contrast images ofonly the bony structures of the joint, but at low resolution. MM-scans provide good overall sofi tissue visualisation. but with very poor resolution. Frozen sections provide hi&-resolution images of the bony structures and sol? tissues ofthe joint, but with poor differentiation behveen similar tissues-
Serial images of milled fiozen specimens can provide hi& resolution images with good soft tissue visualisation (Figure 3.1 ) [16]. However, handling and sectioning frozen specimens requires specialised equipment to ensure that the specimen remains f?ozen throughout the process, and that a satisfactorily cut surface is obtained. Also, milling of frozen specimens unavoidably destroys the specimen. Additionally, images of the cut surface are not mily two-dimensional, as the underlying structures are visible, makhg
S .J. Ferguson Queen's University Biomechanics of the Acetabular Labmm evaluation more ditficult. Finally, it is not possible to differentiate between similar soi? tissue structures, such as the glenoid labrum and adjacent articular cartilage. the meniscus and the articular surraces of the knee, or the acetabular cartilage layers and the acetabular labmm, due to their similar appearance on the digital images. Ideally, digital images should be obtained using a method, which enhances the contrast between neighbouring soft tissue structures.
Polymer embedded specimens are routinely used in histological and gross anatomy studies to investigate the morpho10,aical properties of anatomical structures
[2,3]. To make accurate morphornetn'c measurements of tissue. it is necessary to understand the effect that specimen processing has on the dimensional distortion of the tissue- There have been numerous studies of the dimensionai changes of particular tissues during histolo,oical preparation, many with contlicting results [4-6,10,13]. For example. small specimens of liver and kidney tissue shrink by up to 17% (linear) lbllowing fixation, dehydration and paraffin embedding [6], while large cancellous bone specimens may shrink by up to 7% 1131, a thickness reduction in cartilage specimens of up to 50% following histological preparation has been reported [5], but others have shown that srnaIl specirnens of cartilage and subchondral bone experienced an area shrinkage of onIy 10% (approxirnately 5% linear shnnkage) following ethanol dehydration [l O]. Block plastination, followed by section staining to enhance tissue contrast, has been proposed as a suitable method to study soft tissue components, such as the laryn,~,in their undistwbed state [4], but it is unclear whether such a procedure would be appropriate for Iarger specimens with a mixture of hard and sofi tissues. [t is to be expected that the dimensional changes of the soft tissues in such specimens would be
S.J. Ferbouson Queen's University Biornechanics of the Acetabular Labm
limited by the reinforcing effect of the underlying bone. Thus, serially sectioned, polyrner
embedded specimens could provide accurate geometncal data for the three-dimensional
reconstruction of hard and sofi tissue anatomy.
The goal of this study was to develop a suitable method for serial specimen
imaging in the special case where clear differentiation of visually simiIar, but
compositionally different, coIlagenous tissues is required for large, whole-joint
specimens. Preparation for imaging should not affect the dimensions of the specimen.
Usina this technique, it would be possible to construct accurate three dimensional models of joints which are composed of several different tissue types, each with their own specific properties and function. This technique is dernonstrated on a specirnen taken
fiorn the acetabular rim, which was chosen for its variety of tissue types and composition.
Materials and Methods
Dimensional Changes
Specimens were cut from the acetabular rims of tsvo fiesh-frozen porcine hip joints using a conventional band saw (Bizerba SPA, Milan, Italy). Ovine and porcine sofi tissue have mechanical properties similar to human tissue [7], and so should demonstrate similar dimensional changes during processing. In addition, the cartilage thickness was similar to that of human specimens. Porcine specimens were selected, as they were readily available fiesh From the Iocal slaughterhouse. Specimens were taken from five different locations around the circumference of the acetabular rim. Irregularities in the cut surface were rernoved by -&ding the specimens on a water-cooled, fine grit rotary
S.J. Ferguson Queen's University Biomechanics of the Acetabular Labrum
minding tabk (Struers, RotoPol-25, Copenhagen, Denmark) to provide a flat, planar C
surface for viewing and measurement. During preparation, the fresh specimens were kept
from drying by rinsing with normal saline. The specimens were approximately 10 x
IOmm, 5mm thick, and included a portion of the articular cartilage surface, the acetabular
labrurn, and the underlying subchondral bone (Figure 3.2).
Figure 3 2: Specimens for embedding and subsequent dimensional measurernents were cut frorn five representative Iocations around the acetabular cim, e-g- from the inferior portion of the acetabulum near the transverse ligament. The specimens contained mived tissue types, each with their own specific properties and hnction.
Two groups of five specimens each were prepared (unfixed, fuced). Five specimens were tixed by immersion in a 4% phosphate but'fered paraforrnaldehyde solution for three days, and then rinsed with water. The five fixed specimens and the five
S.J. Ferguson Queen's University Biomechanics of the Acetabuiar Labnun
fresh specimens were dehydrated by immersion in graded aqueous ethanol solutions of
40%, 80%, 96% and 100% ethanol for three days each respectively. The specimens were then embedded following immersion in xylenol for three days, pure methylrnethacrylate
(MMA) for three days, MMA + 2wt% benzoyl peroxide (b.p.) at 5°C for three days, and
MMA + 20~01%plasticiser + 3wt% b.p. at room temperature until the MMA was fully polymerked. Excess MMA above the specimen surface \vas removed using the band saw and the rotary grinding table.
Four linear measurements and one area measurement of each specimen were taken afier each stage of specimen preparation, normalised to the initial measurements of the fkesh specimens. Wet specimens were rneasured while irnmersed in the appropnate preparation medium, i-e. saline or ethanol, to prevent drying. The surface of each specimen was viewed under a light microscope (WILD 308700, Heerbrugg, Switzerland) at a magni fication of xj2. The end points of linear measurements, and the contours of area measurements. were recorded using a digitising tablet that projected the tablet cunor ont0 the specimen in the microscope's field of view (Digikon 4, Kontron Electronics,
Munich, Germany). Three linear dimensions of the sot? tissue were recorded for each specimen using easily identitied landmarks on the specimens for reference (at least one thickness rneasurement perpendicular to the subchondral bone and one tangential measurement alons the articular surface). One linear dimension of the bone was recorded. To quanti* the two-dimensional changes in specimen size, the area of a portion of the soft tissue was measured. Typical measurement locations are shown in
Figure 3.3. Each dimension was measured five times to evaluate erron caused by resolution limitations and operator error, and the average value was used for the
SJ. Ferguson Queen's University Biomechanics of the Acetabular Labrurn calcu1ation of dimensional changes. This series of dimension measurements was repeated after each stage of preparation (5 specimens per group, 5 measurement locations per specimen, 5 repeated measurements per location). Al1 fresh specirnens were measured prior to treatment. Specimens fixed with paraformddehyde were measured after fixation, after dehydration and after embedding in MMA. The unfixed specirnens were measured afier dehydration and atier embedding. Out-of-plane dimension changes were assumed to be the same as the measured in-plane dimension changes.
Figure 3.3: Three linear soi? tissue dimensions were recorded tbr each specimen: at least one cartiIage thickness measurement perpendicular to the subchondral bone (a), one tangential measurernent along, the articular cartilage surface (b) and one other easily reproduced measurernent within the labrum (c). One Iinear dimension of the subchondraI bone was recorded (d), as was the area of soft tissue enclosed by a recognisable and reproducible boundq.
S.J. Ferguson Queen's University Biornechanics of the Acetabular Labrum
Surface Extraction
A 3Ox30x30mm portion of an ovine acetabuIar nm was prepared as descnbed above. The porcine specimens used for the measurernent of dimensional changes were too small for rneaningfiil3D reconstruction. An ovine specirnen was chosen to demonstrate the technique, as future projects in our lab wilI use a sheep model. Following embedding in methyl rnethacrylate, the resin block containing the specimen was ground to forrn two perpendicular edges for later ali-ment dunng image acquisition. The specimen was then sectioned serially, perpendicutar to these two reference edges, using a diamond hole saw (Leica 1600, Leica Instruments GmbH, Nussloch, Germany), to produce 700pm thick parallel slices 1 mm apart. Each slice \vas numbered, and then stained with light green, fuchsin and toluidine blue to enhance contrast between the bone, cartiIage and fibrocartilage. As a staining control, several thin (6 pm) microtome sections of the same specimen were processed with a modified Movat's pentachrome stain, an excellent but more laborious diflerential staining technique for undecakified, plastic embedded thin sections [17]. The pentachrome technique is an established method which stains each tissue type with a visually distinct colour. This provided a clear reference for tissue boundaries when evaluating the quality of the proposed staining technique. The stained specirnens were compared qualitatively (visually) for clear and accurate demarcation of tissue boundaries and colour contrast between tissue types.
Each slice was placed in an alignment fiame and an image taken with a CCD digital co1our carnera (Sony DKC-ID 1, Sony AG, Schlieren, Switzerland), 50mm lens, with a 24-bit colour depth and at a resolution of 768x568 pixels, or approximately
O.OSrndpixe1. Using the SuRFdriver software 1151, contours were manually
S.J. Ferguson Queen's University Biomechanics of the AcetabuIar Labm superimposed over the soft-tissue structures - cartilage and labrum - and the underlying bone, following visible lines of colour and tissue morpho logy separation. These contours were then joined automaticalIy by the software, following identification by the operator of reference aIignment markets in each frame, to form a three-dimensional surface, which could then exported to a commercial FEA sohvare package, Le. 1-DEAS Master Series 5
(SDRC, Milford, USA).
During the measurements, some tissue expansion and shrinkage were noticed.
Considenng linear dimension measurements, the unfixed soft tissue expanded, on average, by 3.5% following alcohol dehydration, and shrunk to 99.7% (+7.8%) ofits original size foIIowing polyrner embedding (Figure 3.4). The fixed soft tissue expanded by 5.4% following fkation, expanded a further 1.4% following dehydration, and shnk to 100.8% (+6.5%) of its original size following polymer embedding (Figure 3.5). The dimensions of the bone tissue remained within i.4% (+L -8%) of their original values for both sets of specimens through al1 preparation steps. There was a large variance in the dimensional changes for ditfèrent measurïng locations. On average, there was a 1.6% standard error in the measurements of each dimension due to operator error or digitising resolution. FoIlowin,o pol ymer embedding, some tissue Iandmarks were not visib te; consequently, the number of measurements for the embedded specimens was lower.
S .J . Ferguson Queen's University Biomechanics of the Acetabular Labrum
dehydraied embedded Processing S tep
Figure 3.3: in-plane dimensional changes of the specimens in the course ofprocessing (unfixed specimens). Error bars indicate * one standard deviation. *polymer embedding obscured some measurement locations.
dehydrateci em bedded Processing Step
Figure 3.5: in-plane dimensional changes of the specimens in the course of processing (formaldehyde fixed specimens), Error bars indicate * one standard deviation. *polymer embedding obscured some measurement locations.
S.J. Ferpson Queen's University Biomechanics ofthe Acetabular Labmm
Staining with light green, fuchsin and toluidine blue enhanced the contrast between different tissue types; the mineralised bone was stained brÏght green, the cartilage blue to violet-blue, and the fibrous connective tissues bIue-green. The fibrocartilage of the
[abmm could be identified as a separate tissue type fkom the articular cartilage by a difference in colour intensity, and the visibly different tissue morphology. The surface staining technique chosen for this study produced comparable tissue differentiation and colour contrast to the pentachrome technique. Tissue boundaries defined by the Li&-
~reen/fuchsin/toluidine-bluetechnique coincided with those shown using the V pentachrome stain. Figure 3.6 shows the digital image of one slice fkom the inferior portion of the acetabular rim, stained to show the articular cartilage surface, the tïbrocartilage at the junction of the acetabular labrum and transverse ligament, and the underlying bone.
Although the software allows automatic contouring using edge detection based on colour threshold values, good results were obtained only when the user manually defined the contour edge points (Figure 3.7) by following visible lines of colour and tissue morphology separation. The final three-dimensional reconstruction of the ovine acetabular rim specimen is shown in Figure 3.8. The finite-element mode1 created with the 1-DEAS tinite element software From the tiled-surface object exported Eorn
SuRFdriver contained approximately 4200 tetrahedral elements. Element resolution was determined by the slice thickness and the number of vertices per contour (approximately
50 vertices per contour).
S.J. Fer-won Queen's University Biomechanics of the Acetabular Labnun
Figure 3.6: A portion of one slice fiom the inferior portion of the ovine acetabular rim. S taining emphasises the contrast between bone (a), cartilage (b) and the fibrocartilage at the junction of the acetabular labnim and the transverse ligament (c). Scale = 10m
Figure 3.7: ManuaI definition of contour vertices, using SuRFdriver sohare, for the 3D reconstruction ofa specirnen cut fi-om the acetabular rim. Tissue morphology and colour were used to determine the boundaries of individual anatomical structures. Scale = I Ornm
S .J . Fer,ouson Queen's University Biomechanics of the Acetabular Labmm
Figure 3.8: The tinal3D reconstruction of a specimen taken fiom an ovine acetabdar rim, showing the fibrocartilage at the transverse ligament/labrum junction (a), the articular cartilage (b) and the subchondral bone (c). The fibrocartilage portion of the mode1 (a) has been meshed with solid elements using 1-DEAS software,
Discussion
The dimensional changes of subsmictures within polymer embedded specimens have been measured. Fixation appears to limit the extent of dimensional changes in the subsequent dehydration step. While others have reported significant shrinkage of soft tissues toIIowing dehydntion and embedding, we found very little effect of the specimen preparation on final specimen size. A one-sample test to detemine statistically significant differences in the dimensions of the embedded specimens relative to the fiesh specimens was not perforrned. Due to the large standard error in the dimensional
S.J. Ferguson Queen's University Biomechanics of the AcetabuIar Labmm
changes for different measuring locations, as indicated by the error bars in Figures 3.3 and 3.4, compared to the small changes in the mean specimen dimensions, this statistical analysis was not necessary. Nevertheless, there was no measurable change in the specimen size after the complete preparation process. Also, as the final dimensions of the embedded specimens were distributed evenly above and below the mean values of
-100%, there is no tendency for tissue shrinkage or expansion. Similady, there were no substantial differences between dimensional changes across the thickness and tangentid to the articulating surface for the soft tissues. It is likely that the underlying bone reinforces the soft tissue and limits any shrinkage or expansion.
While other methods have been developed to generate cross-sectionai images of anatomical structures for three-dimensional reconstruction, the method described here has certain advantages. Surîàce staining the specimens with light green, füchsin and toluidine blue to enhance tissue contrast provided a means to identify the boudaries between anatomicaI structures that othenvise could not be seen. This is important when developing modeis of joints which contain visually similar, but compositionally difièrent, tissues such as the acetabular labrum and acetabular articular cartilage- For undecalcified, methacrylate embedded specimens, the combination of Iight green and tùchsin, two components of the standard Goldner's trichrome stain, stains mineralised bone green and also clearly defines other tissue components, e-g. collagen, muscle and epithelia. However, the staining of cartilage is irregular and non-specific with these two, and the addition of a counterstain is required. Toluidine blue is a standard universal stain, staining non-mineralised and rninerdised cartilage different shades of blue, while having virtually no effect on calcified bone. These two staining techniques are, therefore,
S J. Ferguson Queen's University Biomechanics of the Acetabular Labrurn
complimentary and their combination produced the desired result for our study. Together
they form a trichrome staining technique that is simple and provides a level of tissue
differentiation comparable to that achieved wïth Movat's pentachrome stain, but with
substantial time savings when processing a large number of embedded sections.
Specimens prepared by serial sectioning provide essentially two-dimensional slices. eliminating some of the ambiguity inherent when viewing the cut surîàce of specimens prepared using a serial milling technique. Also, using a serial sectioning process allows the specimens to be kept for further study, which is not possible with a senal miIling technique as the specimen is destroyed during processing. The resolution of the images obtained using this method is lirnited by the choice of digital camera; current cameras ofTer high resolution at a reasonable cost- An alternative method to obtain higher resolution images, at the expense of further processing time, would be to photograph each image with a standard film camera using a fine-grain film. and then scan each individual p hotopph with a flat-bed scanner.
The colour threshold criteria used by the SuRFdriver sofhvare did not always accurately locate the boundaries between structures. The goal of this study, however, was not to specie one particular sohvare package and se-mentation technique, but rather to develop a technique for clearly differentiating visually similar mixed tissue types, and then to dernonstrate a representative anatomical reconstruction- The clearly visible colour and morphology differences between tissue types made it easy to manually define smooth contours. The SURFdriver software is presented here as an option for volume se-mentation and rendering- It is readily available, simple to use and economical. The additional effort of manual contour correction is offset by these factors. More advanced
S .J. Ferguson Queen's University Biomechanics of the Acetabular Labrum
software packages possessing increasingly sophisticated thresholding algorithms could take advantage of the obvious colour contrast between tissue types, such that direct voxel
meshing would bs possible. The individual contours defined by the user were joined by the software to form a tiled surface representation of the three-dimensional geometry, which could be exported in a standard DXF or lGES format and read into a commercia1
FEA package, i.e. 1-DEAS Master Series 5. Conversion of the tiIed surface geometry to a single volume could be performed automatically in 1-DEAS using the option to "stitch" al1 surface edges together, with only minor manual correction required by the user. The ceneration of a tiled surface geometry within SuRFdriver produces a shape visually C faithfùl to the original, but lirnits the possibility of mesh refinement within the FE sofhvare. Alternatively, individual contours can be exported from the se-gnentation software in IGES format, then used to form a solid volume by a "lofiing" operation within the FE software, where a skin is drawn over the three-dimensional space detlned by the cross-sectional contours. This method produces a final solid volume that may not follow the exact shape of the original, due to smoothing, but which allows much more flexibility in the meshing operation. This method has since been applied to the reconstruction of an entire human hip joint, including the cartilage layers and bony structures of the acetabulum and femur, and the acetabular labrum (Appendix A).
A method has been presented for three dimensional shape reconstruction based on digital images of polymer embedded, serially sectioned anatomical structures. Larse preparations containing bone and cartilaginous tissues are dimensionally stable throughout tixation, dehydration and embedding, with linear shrinkage or expansion of less than 2% on average. Before using this method with other tissue types, the
S.J. Ferguson Queen's University Biomechanics OFthe AcetabuIar Labnirn
dimensional changes wouId have to be re-evaluatcd for each specific tissue type. The
additional labour required for this rnethod is offset by several advantages: hi& resolution
images are obtained, specimens are preserved for Wher study, and staining can be used
to enhance tissue contrast and identiQ anatomical features that could othenvise not be
differentiated from their surroundings. This method allows a level of in-plane spatial
resolution and tissue demarcation not possible with some other imaging techniques. such
as CT or MRI scanning, and indeed could serve as a reference standard for models
created with these other techniques,
Acknowledgements
The authors wish to thank the tOIlowing individuals for their contributions to this
study: 1. Keller for assistance with histologica1 specimen preparation and LP. Menfor
iIlustrations (Figures 3 2 and 3 -3).
S.J. Ferguson Queen's University Biomechanics of the Acetabular Labrum
References
[Il Aritan S, Dabnichki P, BartIett R: Program for generation of three-dimensional tinite element mesh from magnetic tesonance imaging scans of hurnan limbs. !tledicd Engineering and Physics 19:68 1-689, 1997
[î] Beck ID, Keaveny TM: A serial-grinding technique for high-cesolution imaging of tra becular bone- Proceedings of the 42nd Annrral Meeting of the Orthopaedic Research Sociew, Atlanta, Georgia, pp. 709, 1996
[3] Dalstn M, Huiskes R, van Eming L: Development and validation of a three-dimensional finite element mode1 of the peIvic bone. Jorrrnal of Biornechanical Engineering 1 17972- 278. 1995
[4] Eckel ME, Sittel C, Walger M, Sprinzi G, Koebke J: Plastination: a new approach to morphological research and instruction with exciscd larynges. Annals of Otol Rlzinologv and laryngology 102:660-665, 1993
[5] Gilmore RStC. Palfiey AS: A histological study of human femoral condylar articular cartilage. J-Anar. 155:77-85, 1987
[6] Iwadare T, Mori H, Ishiguro K, Takeishi M: Dimensional changes of tissues in the course of processing. Jorrri7al of Microscop_t.1 36:323-337, 1 984
[7] Joshi MD. Suh J-K. Mami T, Woo SL-Y: Interspecies variation ot'compressive biornechanical properties of the meniscus. Journal of Biomedical Materials Research 29:893-528, 1995
[8] Kang YK. Park HC, Youm Y, Lee IK, Ahn MH, Ihn JC: Three dimensional shape reconstruction and finite element analysis of tèrnur before and afier the cementless type of total hip replacement. Jotrrnnl of Biomedical Engineering 15:497-504, 1993
[9] Kaplan PA, Bryans KC, Davick JP, Otte M. Stinson WW, Dussault RG: LMR haging of the normal shoulder: variants and pitfalls. Radiofogy 183519-524, 1992
[ 1O] Kaab MJ, Notzli HP, Clark J, ap Gwyn I: Dimensional changes of articular cartilage during immersion-freezing and freeze-substitution for scanning eIectron microscopy. Scanning :LIicroscop_r:International, 1998
[il] Keyak JH. Fourkas MG, Meagher JM, Skinner HB: Validation of an automated rnethod of three-dimensional finite elernent rnodelling of bone. Jorrrnal of Biomedical Engiureering 15505-509, 1993
[II] Kuhn JL, Goldstein SA, Feldkamp LA, Goulet RW, Jesion G: Evaluation ofa rnicrocomputed tomography system to study trabecular bone structure. Jorrrnal of Orrhopaedic Research 82333-842, 1990
[ 131 Lane J, Ralis ZA: Changes in dimensions of large cancellous bone specirnens during histological preparation as measured on slabs from hurnar femoral heads. Cdicified Tissue International 34: 1-4, 1983
S .J . Ferguson Queen's University Biomechanics of the Acetabular Labrum
Merz B. Niederer P, Miiller R, Riiegsegger P: Automated finite element analysis of excised human femora based on precision-QCT, Journal of Bionrec/mrtical Engineering 1 18387- 390, 1996
Moody, D. and Lozanoff, S: SuRFdriver, 1999 (http://www.surf~ver.com)
NLM: The Visible Human Project. National Library of Medicine, 1999 (http://www.nlm.nih.gov/research~visibldvisib1e~human.h~1)
Olah AJ, Simon A, Gaudy M, Hermann W, Schenk RK: Differential staining of calcified tissues in plastic embedded microtome sections by a modification of Movat's pentachrome stain. Staining Technology 52331-33 7, 1977
Pfleiderer M: [;Micrornorion of cementless acetabular arps in the pelvis] (Gerrnan), Ph,D. Thesis, University of Hamburg-tiarburg 1997
Wu JSS, Chen JH: Clarification of the mechanical behaviour of spinal motion segments through a three-dimensional poroelastic mixed finite element rnodel, Medical Engineering and P~~vsics1 8 :2 2 5-324, 1996
S .J. Ferguson Queen's University Chapter Four
The Material Properties of the
Bovine Acetabular Labrum
S.J. ~er~usonl-',J.T. ~r~anr'and K. ltol.'
1A0 ASIF Research Institute, Davos Platz, Switzerland
'~e~artmentof Orthopaedic Surgery, University of Berne, inselspiral, Berne, Switzerland
3 Department of Mechanical Engneering, Queen's University, Kingston, Canada
Submitted to: JozrrnaZ of Orthopaedic Research (1999) Biomechanics of the Acetabular Labnim
Abstract
The compressive and tensile material properties of the bovine acetabular labrum
were measured. Conf-ïned compression testing was used to determine the aggregate compressive modulus and the permeability of the labrum. The compressive modulus of the labnim (0.157 + 0.057 MPa) is comparabie to that of the rnorphologica1Iy similar meniscus, and approximately one-quarter to one-half that of the adjoininz acetabular cartilage. The permeability of the labrum (4.98 f 3-43 x 10-l6 m4/N.s) was lower than that ofthe meniscus and cartilage, with a significantly higher resistance to interstitial fluid flow across the acetabular rirn than along the rim. Specimens from the posterior and superior resions of the labrum were tested to Failure in uniaxial tension. The maximum stress at failure ( I 1-9 -f 6.1 MPa), maximum strain at failure (26.5 t 7.6 %) and tangent modulus (74.7 I44.3 MPa) were similar to those reported For the bovine rneniscus, and to other tissues cornposed of highly-oriented collagen fibre bundles. In tension, the iabrum is much stiffer ( 10 - 15x) than the adjoining articular cartilage, and the posterior region of the labnirn is significantly stiffer (45%) than the superior region. The labrum's low permeability may contribute to sealing of the hip joint. The hi& circumferential tensile stiffness of the labrum, together with its ring structure, reinforce the acetabular rirn and may contribute to joint stability.
Keywords: labrum, acetabulurn, compression, tension, material properties
S.J. Ferguson Queen's University Biomechanics of the AcetabuIar Labrum
Introduction
The acetabular labrum is a fibrocartilaginous lip that is attached to the bony
acetabular rim and deepens the acetabular cup- It is triangular in cross-section and is
attached at its base to the acetabulum, blending smoothly with the articular surfàce of the
acetabuIar cartilage iayer. Its apex is its free margin, which is tumed in against the
femoral head [IO]. The labrum dekes the posterior, superior and anterior boundaries of
the acetabular nm, joining smoothly with the transverse acetabdar ligament to
completely encircle the acetabulum (Figure 4. L).
Figure 4.1 : Media1 view of the bovine acetabulum showing the acetabular labrum (L), the articular cartilage (AC) and the transverse acetabular Iabrurn (TAL). A hurnan pelvis is shown for cornparison.
S J . Ferguson Queen's University Biomechanics of the Acetabular Labm
The most superticial layer of the labrum is a thin membrane characterised by a fine woven mesh of collagen fibrils. Immediately below this is a layer, 20 - 100 prn in ttrickness, with a stratiforrn structure. The third, or inner layer, forming the majority of the labrurn structure, is composed of thick collagen fibre bundles principally arranged parallel to the acetabular rim, with some fibre bundles scattered throughout this layer ninning obliquely to the predominant fibre orientation [24]. It is possible that the smaller, oblique fibres within the body of the labrum serve to tie the larger fibre bundles together, reinforcing the structure of the Iabrum. Shibutani [24] has proposed, based on his morphological studies, that the labrum, with its thick collagen fibre bundles, resists stretching around the circumference of the acetabulum but may be vulnerable to shearing forces across the acetabular rirn, due to the low number of radial fibers,
It is plausible that the labrum is uniquely adapted to its fünction in the hip joint, and that any damage to the labrum rnay contribute to joint degeneration. Several clinical studies have shown that labral injuries, such as tears or intra-articular impingement, lead to changes consistent with osteoarthrosis [1,7,12,13,18]. However, there is little information available about possible mechanisms of labral injury, the role of the labnim in normal hip joint fùnction. or how these are influenced by the labrum's intrinsic structural properties.
To date, the matenal properties of the acetabular labrurn have not been investigated. The labrum, with its highly oriented structure and high water content, is apt to demonstrate a complex, anisotropic behaviour. There have been several studies reported on the matenal properties of the morphologically similar knee menisci
[6,9,1 1.14,20,29] and glenoid labrum [5,22], and one could infer fiom these studies
S .J. Fer,ouson Queen's University Biomechanics of the Acetabular Labrum
representative values for the properties of the acetabular labrum. However, in light of the
increasing clinical interest in the causes and consequences of labrum injury and
pathology, a more thorough understanding of the inûinsic materiai properties of the acetabular labrum is necessary. The objective of this study, therefore, was to evaluate the mechanical behaviour of the bovine acetabular labrum. A specifically chosen subset of mechanical properties were measured in a controlled environment. The properties to be determined in this study w-ere: tissue water content, aggregate compressive modulus, tissue pemeability, matenal parameters describing the non-linear toe region of the tensile stress-strain curve, strain lirnits of the near-linear region of the stress-strain curve, tensile
Young's modulus (fiom the near-linear region), tensile yield strain, maximum tensile strain and maximum stress at tissue failure. Variation of these properties with location or orientation was also eva!uated.
Materials and Methods
Specimen Preparation
Bovine hemi-pelves were obtained fiom a local abattoir within 24 hours of slaughter. Each hip joint was dissected to the level of the acetabular labrum, carefklly disarticulated and then esamined. Any joint with gross evidence of injury or degeneration of the labrum and articular cartilage surfaces was excluded from the study.
Each labrum was separated From the bony margin of the acetabulum using a scalpel.
Individual labrum specimens were wrapped in gauze moistened with a 0.9% phosphate
S J, Ferguson Queen's University Biomechanics of the Acetabular Labrum buffered saline (PBS) solution (pH 7.4), placed in an airtight container and stored at -
20°C until ready for further specimen preparation.
Circular discs of iabrum tissue were prepared for confmed compression testing
To obtain data on the anisotropic permeability of the tissue, entire frozen labrum specimens were sectioned, using a special razor cutting die with parallel blades, either parallel to or perpendicular to the predominant fibre orientation of the labrum. These sections, approsimately 2mm thick, were then mounted on the working stage of a drill press. A sharply tapered hollow coring tool (inside diameter of 3-20 mm) was used to remove a cylindrical plug fiom the partially frozen tissue (Figure 4.2). During the coring operation, a slow cutting speed was used and the sample \vas bathed in a Stream of chilled
(5°C) Ringer's solution to prevent dehydration and heat damage. From the same sections, specimens were cut with a scalpel for the determination of tissue water content.
The specimens were stored in airtight containers at -20°C in gauze soaked with 0.9%
PB S solution until ready for mechanical testing. Previous studies have demonstrated that tieezing does not have an efkct on the material properties of similar tissues [ lb, 15b]. Ten cylindrical specimens for compression testing were hmested for each orientation fiom randomly seiected labrum sections (total specimens n = 20)- Twelve additional tissue sarnples were cut for each orientation for the measurement of tissue water content (total specimens n = 24).
S-J. Ferguson Quren's University Biomechanics of the Acetabular Labrum
1 apex of labrum
bony acetabular margin
Figure 4.2: Preparation of compression testing specimens. 3mm sections were taken pardel to (a) and perpcndicular to (b) the predominant fibre orientation direction ofthe labrum. 32mm diameter cylindrical cores were taken fiom randornly selected labrwn sections-
Entire labrum specimens were sectioned into two regions (superior and posterior
- the anterior portion of the bovine labium did not provide enough material for specimen preparation) by a radial scalpel cut. Each specimen was then straightened and mounted on the freezing stage of a cryotome (Microm HMSOOAG, Car1 Zeiss AG, Walldort:
Gerrnany) with the apex of the triangular labrum oriented approximately parallel to the cutting plane. The apex of the labnim was trimrned in 40 pm increments until a flat cutting surface more than 2mm wide was obtained. Subsequently, 320 pm thick sections were serially microtomed. A thickness of 320 pm was chosen to preserve a large number of collagen fibre bundles. Test specimens were prepared tiom these slices by cutting uniform rectanpular samples using a scalpel (Figure 4.3). Using this technique, the continuity of fibre bundles was preserved, which was not possible using a die-punching
S.J. Fer,ouson Queen's University Biomechanics of the Acetabular Labrurn technique. Fibre integrity \vas verified using a microscope in the subsequent dimensional measuring step. Specimens were obtained fiom two animals and fiom two regions of the labrum (total sarnples n = 64).
Figure 4.3: Preparation of tensile testing specimens. Serial microtome sections 320 pm thick were taken parallei to the apex of the iabrum, Rectangular specimens approximately 2-3 mm wide were cut from the microtome sections. preserving the collagen fibre bundles aIong the Iength of the specimens,
Dimensions[ Measurernents
Prior to testing, the thickness of each compression specimen kvas measured with digital callipers (*O-00 1 mm) viewed under an optical stereo microscope at a magnification of 32x. Measurements were repeated three times for each specimen and averaged. The width of each tensile specimen was measured by viewing through a Nikon profile projector with a calibrated graticule (k0.005 mm) at a mapification of 20x.
W idth measurements were repeated at three equal ly spaced intervals and averaged.
Back-lightinç of the specimen allowed viewing of the fibre structure of the labium to ensure fibre continuity throughout its length.
S.J. Feraauson Queen's University Biomechanics of the Acetabular Labrum
Mechanical Testing
A custom-built mechanical testing device was used to determine the compressive stitiness of the solid rnatrix of labrum tissue and the permeability of the tissue to interstitial fluid flow. Cylindrical plugs of labrum tissue (3.2 mm diarneter) were thawed and rinsed with 0.9% PBS then placed into the 3.3 mm diameter, PBS-filled well of an impermeabte stainless steel confining chamber. The bottom of the well was formed by the flat face of an ultrasound contact transducer (Mode1 V3 16SU, Panametrics, Waltham,
MA, U.S.A.). The plug was loaded from above through a 3.18 mm diarneter porous stainless steel piston (Mott Industrial, Farmington, CT, U.S.A., -20 pm pores, 50% porosity) which had been polished with 1 000 gît polishing paper to provide a smooth. flat contacting surface. The cross-sectional area of the piston was 93% that of the well and 99% that of the labrum specirnen. The porous piston had a pemeability several orden of magnitude higher than that of the labrum tissue. This piston was confiected to the loading platform by a fluted rod. The impermeable bottom and sides of the well ensured that displacement and fluid tlow in the plug was one-dimensional. The entire consolidation charnber was immersed in a 37 +I°C water bath fsee Figure 4.4)-
A tare Ioad of OZON was applied to the loading platform (an average compressive stress of -27kPa) and the system was allowed to equilibriate. The tare load ensured that the labrum specimen filled the confining chamber and that there was good contact to the porous piston. Foltowing equilibtium, a tùrther 0.144N (an average compressive stress of -17kPa) was applied by smoothly lowering a weight with a
S.J. Ferguson Queen's University Biomechanics of the Acetabular Labrum
------motorised linear actuator. The load kvas rnaintained for 36,000 seconds. The displacement of the interface between the porous piston and the labrum plug kvas continuously measured with an ultrasound transducer driven by an ultrasound analyser
(model 5600, Panarnetrics Inc., Waitham, MA, USA) at a repeat rate of 500 Hz. The reflection traces were captured on a digital oscilloscope (mode1 4072, Gould instrument
Systems, Essex, England) and stored on a personal cornputer. A sigature of the reflection pnor to loading was stored as an average of 128 traces. The correlated receiver technique was used to find the time shift between the signature and the current reflected signal, The surface displacement of the plug was then calculated by:
where Ab is twice the time of travel between the transducer and the plug surface at preload (t < O), Ats is twice the time of travel at displacement 6 and c is the speed of sound, corrected for tissue strain [BI. The speed of sound in swollen labrum plugs had been previously determined by applying a range of step strains, via a micron-resolution
Iinear actuator (model VP3 0-25, Newport/M icro-Controle S.A., Evry, France), to the tabmm specixnens. The speed of sound was found to be constant through labrurn tissue for strains in the range of 0-50%, therefore no correction of the speed was required for the strain ranges tested in this study. This confined consolidation technique offers a non- contact displacement measurement free of stiction with a resotution of - I Pm.
S.J. Ferguson Queen's University Biomechanics of the Acetabular Labrurn
Figre 4.4: Confined compression testing apparatus. The cylindrical labmm plug ( 1 ) is seated in an impermeable confuiing chamber (2). A step Ioad, weights placed by a stepper motor (3), is applied fiom above through a porous sintered metal piston (4)- The creep displacement of the [abmm plug top surface is measured using an dtrasound transducer (5). The confining chamber is filled with PBS solution and surrounded by a 37OC water bath (6).
S.J. Fermouson Queen's University Biomechanics of the Acetabular Labm
Due to its hydrated nature, labrum tissue can be modelled as a pomelastic material, with interstitial fiuid moving through, and resisted by, an elastic solid skeleton.
For one-dimensional consoiidation, the displacement of the tissue surFace in the consolidation chamber can be derived from poroelastic theory [4]:
where:
is the time constant of the tissue specimen with a permeability k, a confined compression modulus (or aggregate modulus) HAand an initial thickness h. The intrinsic contined compressive modulus of the tissue's solid skeleton can be determined by the equilibrium displacement of the tissue surface, 60:
At equilibrium, thid exudation fiom the tissue ceases and the applied stress QI is camed entirely by the solid rnatrix of the tissue. With some uncertainty in the equilibrium displacement, the compression modulus HA and the permeability coeff~cient
S.J. Ferbouson Queen's University Biomechanics of the AcetabuIar Labrum
k were determined numericaily using a least-squares criterion, two-parameter non-linear curve fit of equation 4.2 to the experimental consolidation data. Differences in properties betsveen specimens fiom the two orientations were checked for significance using the 2- sarnple Studena's t-test (p < 0.05), following a Shapiro-Wilk's test for norrnality and
Levene's test for equality of variance.
Tende testing was perforrned using a custom-made apparatus mounted in a
uniaxial testimg fiame (mode1 4302, Instron, Canton, MA, USA), Rectangular labrum
specimens were mounted between NO spring-loaded clamps, One clamp was rigidly
Fixed inside an environmenta1 chamber mounted to the stationary base of the testing
machine, and the other was fixed, via a universal joint, to a load ce11 mounted on the
movable crosshead. The range of the Ioad cell \vas 0-IOON with a tested resoIution of
0.04N. The emvironmental chamber consisted of a PBS filled inner chamber containing
the specimen, surrounded by an outer annulus filled with a temperature controlled
circuiating water bath (see Figure 4.5)- A thermocouple in the PBS solution was used to
control the specimen temperature to 37 + 1°C. Following alignment and clampins, the
specimens bvere marked with ink at the edges of the clamps to allow monitoring of slippage. The specimen was preconditioned by rnanually applying a preload of O. IN, and
then allowed to stress relax for two minutes. Followinp relaxation, the specimen was stretched to faillure at a constant rate of O.Srnm/min, or a typical strain rate of 0.001 d,to ensure quasi-static test conditions. Data fiom specimens that slipped or which failed at or
near the grips were discarded.
S.J. Ferguson Queen's University B iomechan ics of the Acetabular Labrum
Fi,pre 4.5: Tende testing apparatus. The rectangdar specimen (1 ) is held between two spring- loaded clamps (2). The lower clamp is fked to the stationary instron testing fiame (3), while the upper clamp is f~vedto a lOON load ce11 mounted on the moving Instron cross-head (4). The specimen is surromded by 37°C PBS solution (5).
S.J. Ferguson Queen's University Biomechanics of the Acetabular Labrum
Load ce11 output and crosshead (and hence grip to grip) displacement were
collected on a personal cornputer at a sampling fiequency of' l Hz. From the load-
displacement data, a stress-strain curve was calculated for each specimen. The strain was
defined as the displacement divided by the original length, equal to (h-1), where h. is
equal to ULo. The stress was defined as the force divided by the initial cross-sectional
area (Lagrangian stress). Specimens were assumed to have a rectangular cross-sectional area, with an area of O.32Omm times the measured width.
The stress-strain cunre for labrum tissue was sigrnoidal, with an initial toe region of increasing stifhess (O < E I&,), followed by a hear region (E,, < & 5 G~~~~),and
finally a yield region of decreasing tangent modulus (E > The tensile stress-strain curve of the tissue was characterised, therefore, with t~vocunre-fitting methods. The first method used linear regression to determine the tensile (Young's) modulus From the near- linear portion of the stress-strain curve between the toe region and the yield point. A yield strain (E~~,~~)was defined as the strain at which the tangent rnodulus (d&) had tàllen below 90% ofits maximum value. The end of the toe region (E~,) was then selected such that the Pearson correlation coefficient for the curve-fit of the near linear region was always greater than 0.995.
The second method for assessing labrum tissue stiffness involved modefling the initial non-linear toe region of the stress-strain curve using an exponential stress-strain law that has been applied to a variety of sol?.tissues, including cartilage, meniscus and ligaments [.i,X ,321.
S.J. Fer-guson Queen's University Biornechanics of the Acetabular Labm
This relationship is given by:
where A and B are rnatrrial coefficients. The derivative, or tangential modulus of the
rnaterial is given by:
Thus, the stifmess of the tissue increases linearly with stress as a result of fibre recruitment as crimped collagen fibres are progressively straightened and carry load. The parameter B represents the rate of change of the tangent modulus with respect to stress, and the product AB represents the tangent modulus of the tissue as the stress approaches zero. The parameters A and B were calculated for each specirnen using a least-square non-linear regession curve-tltting procedure with the solver fèature of Excel (Excel97,
Microsoft Corp., Seattle. USA).
Additional parameters calcuIated for each specimen were the maximum stress
(G,) and ma'rirnum strain (h,) at specirnen failure. A representative curve-fit of the tensile stress-strain curve for a labrum specimen is shown in Figure 4.6. Analysis of variance was used to determine the sigificance of the differences in parameters with respect to location, independent of differences resulting From specimen origin (animal).
S-J. Fermouson Queen's University Biomechanics of the Acetabular Labrum
experiment exponential fit linear fit
strain (mm/mm)
Figure 4.6: GraphicaI ilIustntion of the curve-fit parameters used to define the tensile stress- strain behaviour of the labrum tissue. Stress-strain data were modelled using an exponential stress-strain Iaw for (O < E < E~,). A linear stress-strain law was used for (E,, < E < Euield)-
Water Content
Each specimen was thawed in PBS solution and allowed to equilibriate for one hour. The specimen was then removed from the solution and weighed on a balance
(Mettler AE260, Mettler-Toledo) at'ter gentIy blotting to remove excess moisture. Each specimen \vas then placed in a vacuum drying oven (Vacuthem, Heraus instruments AG,
Zürich) at 65 + 1°C and the oven was evacuated twice to an absolute pressure of 2 x IO-' mbar. Following drying. the specimen was immediately weighed and its water content was determined From the change in weight divided by the original weight.
S.J. Ferguson Queen's University Biornechanics of the Acetabular Labrum
Results
Compressive Material Properties
The compressive properties of the acetabular labrum were measured with samples harvested parallel to and perpendicular to the predominant fibre orientation direction. A representative creep consolidation curve is plotted in Figure 4.7. Most specimens required over 30,000 seconds to reach equilibrium displacement. The mean intrinsic compressive modulus for al1 specimens tested was 0.157 f 0.057 MPa. No statistically sipifkant variations in modulus were found with respect to the orientation of the specimens. The mean permeability coefficient of al1 labrum specimens tested was 4.98 k
3 -43 x 10-l6 rn4/N-S. The permeability coefficient of specirnens harvested perpendicular to major fibre bundles (k = 3.09 + 1.86 x10-l6 rn4/bJ-s) was signiticantly lower (p < 0.0 1) than that of specimens harvested parallel to major fibre bundlrs (k = 6.87 + 3 -66 x 10-I6 m4/N-s). nlat is, there is a Iiigher resistance to the flow of interstitial fluid across the labrum. The mean water content for al1 specimens tested was 72.6 f 3.2%. The results for the compressive properties of the labrum are summarised in Table 4.1.
S .J . Ferguson Queen's University Biornechanics of the Acetabular Labrum
O experiment 1 -aeory
10000 20000 30000 40000 time (sec)
Figure 4.7: Esarnple of experimental compressive creep consolidation data compared with the theoretical prediction based on a poroelastic model. There is good agreement between experimental data and analytical fit for initial slopes and equilibrium displacements.
Tabie 4.1 : Compressive Properties of the Labrum
S pecimen Aggregate Modulus Permeabil ity Water Content Orientation (MW ( 1 O-l6 m4m-s ("4
mean + SE (* SD for water content) * parallel > perpendicular (p < 0.0 1 )
S.J. Ferguson Queen's University Biomechanics of the Acetabular Labrurn
Tensile Material Properties
Typical stress-strain curves for Iabrum specimens are show in Figure 4.8.
Failure of the labrum tissue typically occurred abmptly following a short yield region of decreasing tangent modulus. Some specimens demonstrated a prolonged post-failure response, with individual fibre bundles breaking and slipping past each other, resulting in a jagged stress-strain curve. AI1 specimens analysed in this study failed in the central portion of the tissue. No slippage was observed visually or in the load-displacernent data.
0.1 0.2 strain (mdmm)
Figure 4.8: Typical tensile stress-strain curves are shown for the bovine acetabutar labrum. Some specirnens exhibited a gradua1 failure (A) as individual collagen fibres broke and were pulled through the remaining intact labrum tissue. Most specimens faited abruptly (B. C).
S.J. Feraouson Queen's University Biomechanics of the AcetabuIar Labrum
The variations in material properties of the tissue with respect to location were examined, Parameters describing the !imits of the stress-strain curves are summarised in
Table 4.2. There were no sipifkant diflerences in the observed strain values with respect to specimen location. On average, the non-linear toe region of the stress-strain curve extended to 10.3 + 3.3%, followed by a nez-linear region up to a yield point at
19.5 + 4.3%, and eventually tissue failure at a maximum strain of 26.5 + 7.6%. The maximum stress at tàilure of specimens taken fiom the posterior portion (o= 14.2 h 6.2
MPa) of the Iabrum was significantly higher (p < 0.001) than that of specirnens from the superior portion of the labrum (a= 9.7 =t 5.7 MPa). The average stress at failure for ail specirnens tested was 1 1.9 + 6.1 MPa .
Table 4.2: Measured Tensile Properties of the Labm
Labmm Toe Strain YieId Strain Maximum Maximum Region (%) (%) Strain Stress (%) (MW
Posterior (n = 32) 10.8 + 3-7 20.4 k 4.4 27.2 k 6.8 14.2 4 6.2*
Supenor (n = 32) 9.7 t 2.6 18.5 t 3.9 25.8 k 8.3 9.7 t 5.2
Mean (n = 64) 10.3 13.3 19.5 t 4.3 26.5 t 7.6 11-9 t6.1 mean + SD * posterior > superior (p < 0.00 1 )
S.J. Ferguson Queen's University Biomechanics of the AcetabuIar Labrum
To further characterise the tensile stress-strain behaviour of the labrurn tissue, parameten were fit to the non-linear toe region and the subsequent near-linear region to evaluate tissue stiffness. The resuIts of the regression analysis are surnmarised in Table
4.3. There was no significant variation with location in the exponential parameters that describe the toe region of the stress-strain curve. On average, the tangent modulus as stress approaches zero was 5.83 f 4.1 1 MPa. In the near-Iinear region of the stress strain curve, the mean Young's modulus of al1 specimens was 74.65 + 44.34 MPa. Specimens from the posterior portion of the labrurn (E = 88 -40 ;t 49.8 1 MPa) were significantly stiffer (p < 0.001) than specimens frorn the superior portions (E = 60.90 * 55.53 MPa). which was also refiected in the difference in maximum stress at failure- However, the differences in modulus observed between specimen locations were of the same magnitude as the di fferences observed between specimen sources (animal) .
Table 3.3: Derived Tensile Properties of the Labrum
Toe Region o = ~(e~'-1)
Labrum A B Region (MP~)
Posterior (n = 32) 0.27 f 0.26 24.25t 7.32 5.71 -t 4.37 88.40 t 49.8 1 *
Superior(n=32) 0.30k0.26 22.64 t 6.87 5.94 t 3 -90 60.90 ,t 55.53
Mean (n = 64) 0.28 f 0.26 23.45 + 7.09 5.83 k4.11 73.65 i44.34 mean + SE * posterior > su~rior(p < 0.00 1)
S J. Ferguson Queen's University BiomecIianics of the Acetabular Labrum
Discussion
This study is the fint to examine the mechanical properties of the acetabular labrum. Ln general, the labrum, a morphologically distinct structure of the hip joint, was
Found to possess unique mechanical properties in comparison to the adjoining acetabular cartilage. The perrneability of the labrum kvas much lower than repoded for bovine cartilage [2,23],with a higher resistance to fluid flow radially across the labrurn than along it. The circum ferential tende sti fhess of the labrum was greater than that reported for the adjoining acetabular cartilage [23,32]. Furthemore, the tensile stiffhess and strength of the labrum was greater in the posterior compared to the superior region.
Study Design
Although the results remain valid, there are some limitations to the study. Bovine labrum tissue \vas selected for testing instead of human tissue. With the growinz interest in the clinical consequences of Labrum injury, it would be desirable to measure the properties of human labrum tissue. However, human tissue is not as readily available for testing as bovine tissue, and the size of the human labrum can make specimen preparation difficult. Joshi et al. [14] demonstrated that there is no statistically significant difference in the mechanical properties of human and bovine rneniscal tissue, a fibrocartilaginous tissue morphologically similar to the labrum. Bovine tissue has ofien been selected for the evaluation of meniscus and cartilage material properties [6.15,19,20,23].
The rectangular tensile specimen geometry and the use of g~p-to-@p displacement measurement for the calculation of strain introduce some uncertainty into
S J. Ferpon Queen's University Biomechanics of the AcerabuIar Labrum the results. However, specimen dimensions were taken as the average of three measurements for each, and specimens with more than a 5% deviation in width from the average measurement were rejected, therefore ensuring a unifonn rectangular cross- section. Specimens which slipped during tensile testing or which failed at or near the grips, were also discarded. Preparation of rectangular specimens instead of dumbbell V specimens ensured continuity of collagen fibre bundles along the length of the specimen.
The low strain rate used in this study (0.00 1 s-') did not allow the determination of the strain-rate dependant behaviour of the tissue- However, the goal of the study was to determine the intrïnsic tensile properties of the collagenous solid-matrix of the tissue, independent of frictional drag effects induced by interstitial fluid motion at hi& strain rates. To achieve this goal, quasi-static test conditions were required.
The large variance in the measured properties is not a resuIt of any limitation in study design, but rather is typical for studies of the mechanicd behaviour of sofi tissues.
Similarly large ranges of vaIues have been reported for the properties oc for example, the meniscus [9,20].
The tensile properties deterrnined in this study cannot adequately describe the compressive behaviour of the tissue, and the compressive properties determined From the creep consolidation testing cannot adequately describe the tensile behaviour of the tissue.
Therefore, care must be taken that the appropriate constitutive relationships are chosen for the behaviour that is being modelled when using these properties in, for example, finite-element code.
S.J. Ferguson Queen's University Biomechanics of the Acetabular Labm
Compression Testing
The compressive behaviour of the Iabrum tissue is best described by a model that
considers the material as a poroelastic sotid, with interstitial fluid contained within the open pores of a penneable collagen-proteoglycan mat+. The behaviour of such biphasic tissues has previously been described using a theoretical model [k9] which has been
subsequently incorporated into finite-element code [26]. It has also been shown that such a tissue can be accurately modelled using a poroelastic model [25], previously used to describe the behaviour O f fl uid-saturated soils. Such a mode1 is available in commercial finite-element packages, and recently Wu et al 1331 demonstrated the validity of solutions obtained From such commercial software.
Such a constitutive model was used to determine the intrinsic compressive modulus of the solid matrix of the labrum and its permeability. With its hi& water content (>70%) and fibrillar collagen nétwork, the tissue was expected to demonstrate a creep response sirnilar to that observed for other hydrated sofi tissues, such as articular cartilage or meniscal tissue. In general, the theoretical creep consolidation curves closely fit the observed experimental data, and were qualitatively similar to the creep responses for cartilage and meniscus. The average intrinsic compressive modulus of the bovine labrum is approsirnately one quarter to one half that reported for bovine articular cartilage [2,19] and is similar to that reported for the bovine rneniscus [ 14,201. The average permeability of the labt-urn tissue was lower than that reported for bovine articular cartilage [2,23], and also lower than that observed in the bovine rneniscus
[14,20]. The low permeabiiity and aggregate modulus of the iabrum greatly influence its creep response, with a correspondingly large time constant, as evidenced by the long time
S.J. Ferguson Queen's University Biomechanics of the Acetabular Labrum
------
(-36,000 sec) required for the tissue to reach its equilibrium displacement. For a
poroelastic solid, the frictional drag forces generated during interstitial fluid flow are
inversely reiated to the tissue's permeability, When the tissue is loaded, energy is
dissipated through these fictional drag forces. it is possible that the low pemeability of the labrum contributes to the shock absorbing capacity of the tissue, minimising the
severity of impacts, especially at the estreme positions of joint rotation, where there is an
increased risk of dislocation. The results of this study also showed that there was a significant difference in the penneability of the labrum tissue parallel to and perpendicular to the predominantly circumferential coilagen fibre bundles. There is a oreater resistance to interstitial fiuid flow across the labrum. Previously, parametric finite Y element analysis has been used to study the influence of the mechanical properties of the labrum on its sealing ability [SI. A high resistance to radial interstitial tluid flow out of the articular cartilage at the periphery of the joint, as measured in this study, decreases the rate ofcartilage layer consolidation and enhance the ability of the cartilage iayers of the hip joint to carry loads through interstitial tluid pressurisation.
Tension Testing
The tensile stress-strain curves for the Iabrurn specimens exhibited a si-moid shape similar to those reported for other soft tissues such as the meniscus, cartilage and ligaments [3,20,39,32]. Most studies have amibuted the initial toe-region of the curve for such tissues to the sequential recruitment and stretching of crimped collagen fibres.
Following the toe-region was a near-Iinear region, where al1 collagen fibres have been recruited and are uniformly extended. Prior to failure, the tangent rnodulus of the stress-
S.J. Ferguson Queen's University Biomechanics of the Acetabular Labrurn
strain curve decreases. as the materiai 'yields". This behaviour has been attributed to the
sequential failure of individual collagen fibres, with fibres failing in the same sequence in
which they were recruited [17]. The final failure of the tissue is abrupt, as can be seen in
the representative stress-strain curves. However, some specimens exhibited a prolonçed
post-Failure response with a gradua1 loss ofstrength, as individual fibre bundles could be
seen to fail and pull through the remaining intact labrum tissue. Although no quantitative
measures of collagen fibre bundle density or size were made, notes were made of the
tissue rnorphology of specirnens based on visual observation during dimension
measurement (under the microscope), and also during testing (unmagnified). Specimens
with the most uniform appearance, and with the majority of fibres arranged along the
length of the specimen, exhibited the highest moduli and tensilr strength. Specimens
with a less organised fibre structure were sofier and failed at lower stress levels. The
former are more representative of the physiological state. The different failure modes are
aIso retlected in the large variance in the results ofthe tensile tests.
The initial tangent modulus of bovine labrum tissue as strain approaches zero,
which is the product of the exponential curve-fit parameters A and B, was 5.8 + 4.1 MPa
on average. This is similar to the average linear (Le. post-toe region) modulus of bovine
articular cartilage [23,32]. The tangent modulus increases through the toe-region. The
rate of change of the modulus is measured by the exponential parameter B. Typical of a tissue whose properties are heavily influenced by its fibrous structure, the modulus of the
labrum increases rapidly with increasing strain. with the value of this parameter B greater on average than, for example, that reported for the fibrous glenohumeral ligament [3].
The modulus calculated for the near-linear portion of the stress-strain curve, 74.7 + 44.3
S J. Ferguson Queen's University Biomechanics of the Acetabular Labrum
MPa on average, was similar to that reportrd for the bovine and human meniscus
[9,20,29,3 11 and, as previously noied, much higher than that of the adjoining articular cartilage [23,32]. The only signiticant differences in mechanical properties with location were differences in modulus and maximum stress (p < 0.00 1) with the posterior region of the labrum stiffer and stronger than the superior region. However, it is important to note that the difference observed in modulus between locations was of the same magnitude as that observed between sources (animais). As the maximum failure strain (26.5 t 7.6%) was similar for both regions, the increased stiffhess and strençth of the posterior region could be explained by a greater density of collagen fibre bundles. A detailed morphological study is required to verify this hypothesis.
The Iabrum is a highly oriented tissue, with the majority of its large collagen fibre bundles running parallel to the acetabular rim. Histological evaluation has demonstrated free nerve endings in the superficiai layers of the Iabrurn [16]. The labrum, together with the transverse acetabular ligament, encircles the femoral head and deepens the acetabular socket beyond half a sphere. Dislocation of the &moral head within the acetabulum would produce considerable circumferentia1 strains within the Iabrum, providing the stimulus for a possible proprioceptive mechanism in the hip joint. The labrum may reinforce the acetabular rim [2 L] and provide passive resistance to joint dislocation, tlirou&h its oriented. fibre-reinforced structure (see also Appendix B for an analysiç of such a stabilising mechanism). This passive resistance would be enhanced by any pre- tensioning of the labrurn in its resting state, The labrum has also been show to enhance joint stability by sealing the joint and providing a negative intra-articular pressure upon joint dislocation [27,30]. It is possible that the fibrous, low-permeability Iabrum, if
S.J. Fermauson Queen's University Biomechanics of the Acetabular Labrum
snugly opposed against the femorai head, could seal the joint against fluid expression
from the intra-articular joint space during normal loading, enhancing joint lubncation
through retention of a lubricating fluid film. Previous analysis has shown that, within the
range of circumferential stiffness determined in this study, the labrum maintains tight
contact with the femoral head, preserving the intep-ty of this seal [g].
This study provided rneasurements of the compressive and tende material
properties of the bovine acetabular labrum. These properties can be used as the basis for
modelling of the function of the acetabuiar labrum to further understand the role of the
labrurn in normal hip joint biomechanics, possible injury mechanisms of the labrum, the
consequences of Iabrurn darnage on long term joint health, and perhaps even the efficacy
of proposed labrum surgicai repair techniques.
Acknowledgements
Thanks are due to 1. Keller for her assistance in the preparation of specimens and
to Dr- P.Pyk for enhancements to the consolidation software.
S.J. Feraouson Queen's University Biomechanics of the Acetabular Labrum
References
Atenberg AR: Acetabular labrurn tears: a cause of hip pain and degenerative arthritis. Sozrtliern Medical Jozirnal70: 1 74- 175, 1977
Arnosczky SP, McDevitt CA, Schmidt MB, Mow VC, Warren RF: The effect of cryoprese~ationon canine menisci: a biochemical, morphologie and biomechanical evaluation. Journal of Orthopaedic Research 6: 1- 1 2,- I 988
Athanasiou KA, Agmal A, Muffoletto A, Dzida FJ, Constantinides G. Clem M: Biomechanical properties ofhip cartilage in experimental animal models. Clinical Orfhopaedics and Related Research 3 1 6:254-266. 1 995
Bigliani LU, Pollock RG, Soslows@ LJ, Flatow EL, Pawluk RJ, Mow VC: Tensile properties of the inferior glenohumeral ligament. Jo~rrnalof Orthopuedic Reseurcfi 1 0: 1 87- 196, 1992
Biot LMA:General theory of three-dirnensional consolidation. JAppl.Phys. 13: 155- 164, 1941
Carey JP R: Compressive Characteristics of rhe Glenoid Labnrrn, M.Sc. Thesis, Queen's University. 1998
Chem KY, Zhu WB, Kelly MA, Mow VC: Anisotropic shear properties of bovine meniscus. Proceedings of the 36th Annrral hfeeting of the Orthopaedic Resear-ch Socie~. New Orleans, Louisiana, pp. 246, 1990
Dorrell JH, Catterall A: The tom acetabutar labmrn. Jorrrnal of Bone andJoinr Srrrgery [w68-B:400-403, 1986
Ferguson SJ. Bryant JT, Ito K: An investigation of the function of the acetabular labmm using a poroelastic finite element model. Jorrrnal of Bone andJoinr Srirgey BI-^ 8 l- B:SUPP 169, 1999
Fithian DC, Kelly MA, Mow VC: Material properties and structure-fllnction relationships in the menisci. Clinical Orthopaeciics and Related Research 252: 19-3 1, 1990
Gru-v's Anu(omy: The Anafornical Buis of Medicine and Srrrgen, 3 8, pp 684. Ed b y PL Williams. New York, Churchitl-Livingstone, 1995
Hacker SA, Woo S L-Y, Wayne JS,Kwan MK: Compressive properties of the human meniscus. Proceedings ofthe 38th Annzral Meetiulg of the Orthopaedic Reseal-ch Sociew, Washington, D.C., pp, 627, 1992
Harris WH, Bourne RB, Oh 1: Intra-articular acetabular labrurn: a possible etiological factor in certain cases of osteoarthritis of the hip. Jozrrnal of Bone andJoint Srrrgery 6 1 - A:510-514, 1979
Ikeda T. Awaya G. Suailci S, Okada Y, Tada H: Tom acetabular labmm in young patients. Artllroscopic diagnosis and treatment. Jorrrnal of Bone andJoint Srrrgery [Br] 70-B: 1 3-1 6, 1988
S.J. Ferguson Queen's University Biomechanics of the Acetabular Labmm
[13] Joshi MD. Suh J-K, Marui T, Woo SL-Y: interspecies variation ~Fcompressive biomechanical properties of the rneniscus. Jotrrnal of Biomenical :Materials Research 295323-828, 1995
[ 151 Jwelin J, Buschmann MD, Hunziker EB: Characterization of the equilibrium response of bovine humera1 cartilase in confined and uncontined compression. Proceedings of the 41st Annzral Meeting of the Orthopaedic Research Sociefy, Orlando, Florida, pp. 5 12, 1995
Cl Sb] Kietèr GN, Sundy K. McAllister D,Shrive NG, Frank CB, Lam T. Schachar NS: The effèct of cryopreservation on the biomechanical behaviour of bovine articular cartilage. Jow-nul of Orrhopaedic Researclî 7:494-50 1, 1989
Kim YT, Amma H: The newe endings of the acetabular labrum. Clinical Orthopaedics and Relared Research 3 20: 1 76-1 8 1, 1995
Liao H, BeIkoff SM: A failure mode1 for ligaments. Jozrrnal of Biornechanics 32:183-1 88, 1999
McCarthy JC, Busconi B: The role of hip arthroscopy in the diagnosis and treatment of hip disease. Canadian Jozlrnal of Surgery 38 Suppl 1:s 13-1 7, 1995
Mow VC, Kuei SC, Lai WM, Armstrong CG: Biphasic creep and stress relavation of articular cartilage in compression: theory and experiments. Jortrnal qf'Biornechanical Engheering 102:73-84, 1980
Proctor CS, Schmidt MB, Whipple RR, Kelly MA, Mow VC: Material properties ofthe norrnaI medial bovine meniscus. Jorrrnal of Orthopaedic Resemch 7:77 1-782. 1989
Putz R, Schrank C: [Anatomy of the Iabrocapsular complex of the hip joint]. Ortitopade 27:675-680, 1998
Reeves B: Experiments on the tensile strength of the anterior capsular structures of the shoulder in man. Jortr-nal oj'Bone crnd Joint Swgev [Br] 50-B:858-868, 1968
Roth V, Mow VC: The intrinsic tensile behavior of the matrïx of bovine articular cartilage and its variation with age. Jortrnal of Bone and Joinr Szrrgery 62-A:1 102-1 1 17, 1980
Shibutani N: Three-dimensional architecture of the acetabular Iabrum - a scanning etectron rnicroscopic study. Jozrrnal of the Jnpanese Orthopaedic Association 62331-329, 1988
Simon BR: MuItiphase poroelastic finite element models for sofl tissue structures. Applied Mechanics Revieir: 45: 19 1-2 18, 1992
Spilker RL, Suh JK. Mow VC: A finite eiement formuIation of the nonlinear biphasic mode1 for articular cartilage and hydrated sofi tissues including straindependent permeability. In: Comptrtational Methods in Bioengineering., 81 -92. Ed by RL Spilker and BR Simon. New York, American Society of Mechanical Engineers, 1988
Takechi H, Nagashima H, Ito S: Intra-articular pressure of the hip joint outside and inside the lirnbus. Jotrrnal of the Japanese Orthopaedic Association 56:529-53 6, 1982
S.J. Ferguson Queen's University Biomechanics of the Acetabular Labnrm
Tepic S :Dynanrics of and entrop-vprodrction in the cartilage luyers of the qnovial joint,, Sc-D. Thesis, Massachusetts Institute of Technology- 1982
Tissakht M, Ahmed AM: Tensile stress-strain characteristics of the human meniscal material, Jorrrnal of Biornechanics 1-422, 1995
Weber W, Weber E: Uber die mechanik der menschlichten gehwerkzeuge nebst der beschreibung eines versuches uber das herausfallen des schenkelkopfes aus der pfanne im hftverdemten raum. Annalen Pfzysik und Chernie 40: 1 - 13, 1837
Whipple RR, Wirth CR, Motv VC: Anisotropic and zona1 variations in the tensile properties of the meniscus. Proceedings of rhe 3Isr Annuul Meeting of the Orthopaedic Resmr-ch Society, Las Vegas, pp. 367, 1985
Woo SL-Y, Akeson W, Jemmott GF: Measurements of nohomogeneous directional mechanical properties OF articular cartilage in tension- Jorrrnal of Biomechanics 9:785-79 1, 1976
Wu JZ, Herzog W, Epstein M: Evaluation of the finite dement sohvare ABAQUS for biomechanical modelling of biphasic tissues. Joirrnal of Biornechanics 3 1 :165-1 69, 1998
S.J. Ferguson Queen's University Chapter Five
The Acetabular Labrum Seal:
a Poroelastic Finite Element Model
S.J. ergu us on'-^, J.T. ~r~ant',R. ~anz'and K. ~to'"
I A0 AStF Research Institute, Davos Platz, Switzerland
'~e~artmentof Orthopaedic Surgery, University of Berne, Inselspital. Berne, Switzerland
3 Department of Mechanical Engineering, Queen's University, Kingston, Canada
In Press: Clinical Biomechanics (2000) Biomechanics of the Acetabular Labrum
Abstract
A finite element mode1 of the acetabular labrum is presented. The model was used to investigate the labrum's ability to seal a pressurised layer of synovial tluid within the joint, and to study the influence of this sealing mechanism on cartilage deformation, interstitial fluid pressure and solid matrix stresses. The model kvas an axisymmetnc geometric approximation of the acetabular and femoral cartilage layers and the C surrounding labrum. A poroelastic formulation was used to account for the solid and fluid components of these hydrated tissues. The highly oriented circumferential tibres of the labrum were represented by transversely isotropic material properties. A sensitivity analysis of the labrum material properties was carried out. The results show that the labrum may play an important role in joint lubrication and load transkr. With a compressive load of 1200 N applied across the joint model, the labrum could seai a layer of prsssurised fluid between the femur and acetabulum if the circumferential stiffness of the labmm was greater than 100 MPa, thus preventing contact of the articulating surfaces.
With this sealing effect, loads were transferred across the joint predominantly by uniform pressurisation of the interstitial tluid of the cartilage layers. In the absence of this sealing, strains and stresses in the solid matrix of the cartilage layers were higher.
Keywords: labrum, hip, cartilage, sealing, finite element
S.J. Ferguson Queen's University Biomechanics of the Acetabular Labrum
Introduction
The acetabular labrurn is a fibrocartila,binous rirn attached to the osseous margin of the acetabulum with a tissue structure sirnilar to the knee menisci. It deepens the acetabular socket and extends the coverage of the femoral head. InferiorIy, it joins smoothly with the transverse acetabular ligament to bridge the acetabular notch, forrning a complete circle. It is triangular in cross-section, with its base attached to the acetabulum and its apex forming the free edge of the labrum, which is tumed in against the femoral head.
Little is known about the acetabular labrum, its significance in normal joint hnction, and the possible consequences of labrum pathology. The labrum is cornposed of hydrated tissue, with collagen fibre bundIes oriented predominantly in the circumferential direction, aligned with the acetabular rim [26]. Like the morphologically simiIar meniscus, it is plausible that the labrum is uniquely adapted to its function in the hip joint, and any damage to the Iabrum may contribute to joint degeneration. Several clinical studies have shown that labrai injuries, such as tears or intra-articular impingement, lead to changes consistent with osteoarthrosis [3,6,12,16,19].
Observations have been reported which are consistent with a sealing function of the labrum. Simple experîments with cadavers were conducted to study the role of the labrum in stabilising the hurnan hip joint [34]. in a more comprehensive study, Takechi et al [29] measured the intra-articular pressure in the hip, using a needle pressure transducer, both inside and outside the labrum. Both groups concluded that the labrum contributed to the stability of the hip joint through a partial seal of the joint, creating a
S.J. Ferguson Queen's University Biomechanics of the Acetabular Labrum negative intra-articular pressure upon joint distraction, and by the structural resistance to dislocation provided by the labrum tissue itself Terayama et al 13 11 demonstrated that a
0.2-0.6 mm thick Ruid film remained sealed bettveen the articulating surfaces of ti-esh cadaverk hip joints afier applying a 1000 - 1500N compressive load across the joint. In these joints, the labrum remained in tib&t contact with the opposing femoral cartilage surfàce (Figure 5.1). A similar sealing and stabilising phenornenon has been demonstrated for the glenohumeral joint [ 13,17,3 31. Besides improving the stability of the joint, this possible sealing hnction of the labrum couId enhance a fluid fiIm lubrication mechanism in the hip joint and prevent direct solid-on-solid cartila,oe contact.
Figure 5-1: Penpheral portion of the hip joint under a Ioad of 1275 N. The acetabular and femoral cartilage surfaces did not corne into direct contact, A joint space of approximately 0-4 mm remained. The joint space bebveen the labrum (L) and the fernord cartilage tapered towards the peripheïy. The articular cartilage of the femoral head was markedly depressed where the labrum covered the fernord head (arrow). Adapted fiom Terayama et al [3 11.
S.J, Ferguson Queen's University Biomechanics of the Acetabular Labm
By sealing against fluid expression fkom the cartilage layers, loads applied to the joint are carried by fluid pressure within the cartilage, shielding the collagenous solid matrix of the cartilage from high stresses [4,18,27]. Failure of this seal would lead to higher loading in the solid matrix of the cartilage surfaces and increased friction, contributing to the degenerative changes of osteoarthrosis. The goal of this study was to investigate the ability of the acetabular labrum to seal the hip joint, using a finite element computer model, and aIso to study the influence of this sealing mechanism on cartilage deformation, fluid pressures and so lid stresses-
Method
To study the ability of the labrum to seal a fluid layer within the hip joint, an a..isymmetric finite element model was designed to approximate the study conducted by
Terayama et al [311. The commercial finite element software package ABAQUS 5.7
[1,2] was chosen for this study, due to its ability to model contact mechanics and poroelastic (biphasic) materials. The model included the articular cartilage layers of the fimur and acetabulurn, and the acetabuIar labrum (Figure 5.2). The cartilage was modelled as an isotropic poroelastic rnaterial. The rnaterial properties chosen were consistent with previous analyses of cartilage biornechanics [24,32,35]: E = 0.467 MPa, v
= 0.167, k =+7.358~10-'mds, specific weight of the pore fluid w.8 1 kN/m3 and solid fraction $,= 20 percent of the total tissue volume. The labrum, with its highly oriented structure of circum ferential collagen fibre bundles, was modelled as a transverse1y isotropic poroelastic material.
S.J. Ferguson Queen's University Biomechanics of the Acetabular Labrum
rigid imperneable
Figure 5.2: Axisymmetric tinite-element mesh of the hip joint including the acetabular and fernoral cartilage layers, acetabular labrum (grey elernents) and a fluid-fiIled intra-articular space. Bone is assumed to be rigid and impermeable. The elements are shown reflected about the axis of ~~vmrnetryfor clarity.
S J . Feraouson Queen's University Biomechanics of the Acetabufar Labmm
The circurnferential sti&ess was varied parametrically over a range from 50 - 200 MPa. based on the properties of the structurally similar meniscus 110,251 and our own pilot tests. It was assurned that the circumferential stiffness would be the most important factor determining the results of the model. The modulus of the cartilage elements was also varied over a range of jO-2OO% of its nominal value to assess the sensitivity of the rnodel solution to cartilage material properties.
Hip joint geometry was based on representative measurements of acetabular and femoral hrad diameter and cartilage layer thickness, taken î?om several clinical MRI data sets of non-symptomatic hip joints (Department of Orthopaedic Surgery, University of
Berne). These were consistent with geometry f?om cryosection images of the Visible
Human data set [22]. The femoral head had a radius of 26mm. the opposing cartilage layers were 3mrn thick and the underlying bone was considered rigid and impermeable.
The joint \vas modelled as slightly incongnient, based on anatomical observation [5,8], with initial contact between the kmur and the labrum at the rim of the acetabulum. This resulted in a maximum initial fluid layer thickness of 0.4 mm behveen the femur and the acetabulum-
The tluid layer and acetabular fossa (fat pad) were modelled using poroelastic elements with low stiffness, hi& permeabiiity and high water content to simuIate spaces filled with an incompressible fluid which could develop a hydrostatic pressure, could be easily redistributed and which could be freely exchanged with the interstitial fluid of the cartilage layers. Interface elements with sliding capabilities were placed along the interface between the acetabulum and the femur. Further details are given in Appendix
D.
S.J. Ferguson Queen's University Biomechanics of the Acetabular Labnim
A compressive ioad of 1200 N was applied along the axis of symmetry and held
for 1000 seconds. The contact pressure behveen the labrum and femur was compared to
the pressure in the fluid layer. If at any time the fluid pressure exceeded the average
contact pressure behveen the labrum and the femur, then the labrurn seal was considered
to have been compromised, and the mode1 solution was ended. For cornparison, a second
mode1 was run without a tluid layer between the articular surfaces, to study the
consequences of direct contact between joint surfaces. Contactinj cartilage surfaces were
assumed to be sealed against fluid flow perpendicular to the surface; non-contacting
surt'aces were assumed to be fiee-draining.
The mode1 demonstrated thar the labrum could seal a layer of pressurised fluid in
the space between the femur and the acetabulum when the stifhess of the circumferential
fibres ofthe labrum was greater than 100 MPa. There was an initial deformation of the
labrum following loading, but the two cartilage sutiaces remained separated by the fluid
layer. The cartilage surfaces gradually approached each other as the fluid redistributed
itself. The incompressible fluid was squeezed into the space at the periphery of the joint created by gradua1 deformation of the cornpliant labnrrn over time. Under this static step
load, the femoral cartilage first came into contact with the acetabulum 200 seconds afier the load was applied (for EciK= 200 MPa). As the fernur \as pressed hto the acetabulum. deformation of the ringstructure of the Iabrum was resisted by the circumferential stiffness of its oriented collagen fibres, and this maintained tight contact
S.J. Fereauson Queen's University Biomechanics ofthe Acetabular Labm
bettveen the labrum and the femur- Peak circurnferential strains in the Iabrum approached
2% (Figure 5.3). The contact pressure between the labrum and femur was constant.
Figure 5.3: Circumferential strain (E33) and stress (S33) in the labrurn resulting fiom the proposed sealing function.
S.J, Feraouson Queen's University Biomechanics of the AcetabuIar Labnrrn
With a fluid layer sealed beîween the two cartilage surfaces, Ioads were transferred through the cartilage layers of the joint predominantly by pressurisation of the entralned interstitial fluid. The fluid pressure within the tossa and the cartilage layers showed a very even, uniform distribution across the joint, with a nominal value of 6 10 kPa (Figure
5.4). Only at the periphery of the joint, where the labmm was in contact with the kmur,
\vas load transferred across the joint through stresses in the cartilage solid rnatrix. As the stiffness ofthe labrum decreased, the whole joint became more cornpliant- The kmur deflected 25% more after the initial load application when the circumferential stiffness was decreased by 50%. if the circumferential stiffness of the labrum was less than 100
MPa, the contact pressure between the labrum and Femur was lower than the hydrostatic pressure driving tluid out ofthe joint space, and hence wouId not provide a seal.
If this fluid layer was not present, there was direct contact bettveen the femoral and acetabular cartilage. The resutting distribution of fluid pressure and strain within the cartiIage layers \vas Iess unifonn (Figure 5.5)' with a higher peak tluid pressure (870 kPa) at the centre of contact, but with lower fluid pressures, and consequently higher solid strains, towards the perimeter of the joint. For example, the principal compressive strains within the acetabular cartiIage adjacent to the labrum were much higher (20% vs. 3%). A larger portion of the appkd load was transferred across the joint through direct soIid on solid contact.
S.J. Feraouson Queen's University Biomechanics of the AcetabuIar Labrum
S. Min. Pnncipa) IAve. Cm: 75%) t1.145601 +5.000e-02 +3.000e-02 +1 .M)(3e-02 -1.000e-02 -3.m2 -5-2 -7.QObQ2 -9.WOe-02 -1.100eQ1 -1.3ooe-01 -?5OOesl -3.446801
Figure 5.4: Interstitial fluid pressure (POR)and principal compressive solid matnu stresses (S-Min-Principal)are plotted imrnediately after load application. The Iabrum sealed a layer of pressurised fluid between the cartilage layers, preventing solid on solid contact. The rnajority of the load that was applied to the hip joint was carried by fluid pressures (a) nther than by cartilage solid stresses (b). Fluid pressure supports up to 95% of the applied load in the cartilage layers. Stresses in the solid matrix are almost zero throughout the acetabular cartilage layer. Pressures and stresses are given in MPa.
S.J. Fereauson Queen's University B iomechanics of the Acetabular Labmm
WR (Ave. Cnt: T5%)
S. Min. Pnncioal IAve. Cric 7541
-1 .a00842 -3.000e-02 -5.COOeo2 -7.C-02 -9.LlOûe-02 -1 .lODe-01 -1.300e-01 -1500e-01 -1.929e-01
Figure 5.5: With no labrum sealing, there was direct contact between the femoral and acetabuIar cartilage. The distribution of fluid pressure (POR) (a) within the cartilage layers was les uniform, with a higher peak fluid pressure at the centre ofcontact. in contrat to Figure 5.3, compressive solid matrix stresses (S.Min.Principa1) (b) deveioped throughout the acetabular cartilage layer. Pressures given in MPa.
S.J. Ferguson Queen's University Biomechanics of the Acetabular Labrum
Discussion
The studies of Weber et ai [34] and Tackechi et ai [29] demonstrated that the labrurn itself could partially seal the hip joint, preventing fluid flow into and out of the joint space as the joint was lwated- These studies revealed a possible function of the labrum, and provided stimulus for further investigation; does this sealing effect exist under a compressive, physiological load? Consistent with the experimental observations of Terayama et al [3 1f, this tinite element analysis showed that the acetabular labrum could seal a pressurised layer of fluid within the joint space of the hip for an appreciable penod of time when the joint was subjected to a compressive load. The labrum's highly oriented circumferential fibres, which resist deformation under the pressure exerted by the fluid in the joint space, and the slightly incongruent geometry of the hip provide this sealing function. Parametric analysis showed that this sealing function was possible over a range of circurnferential stif'hess values sirnilar to those reported For the morphologically similar meniscus (90 - 300 MPa) [IO].
In their study, Terayama et ai [3 11 injected micron-sized ink molecules in solution into the joint space as a fluid rnarker. The experimentally observed motion of fluid tiom the joint space towards the periphery of the joint was also seen in this finite element analysis. The maintenance of contact pressures, predicted by the cornputer model, behveen the Iabrum and the femoral cartilage, which provided this sealing fùnction, may explain the marked depression observed in the expenments where the labrum contacted the femoral cartilage. Contact pressures predicted behveen the labrum and femur were uniform across the width of the labrum. Therefore, consideration of the average contact
SJ. Feraouson Queen's University Biomechanics of the Acetabular Labrurn
pressure as a criteria tor determining the sealing performance of the labmm would seem to be justified by this observation. However, the possibility exists to eventually develop a
more sophisticated model, which could simulate thin-film flow through this contact
region (a "leaky" seal). The fluid space observed by Terayarna et al [3 11 \vas similar to that predicted by the model, and was much greater than the squeeze film thickness predicted in theoretical analyses of hip joint lubrication [7,9,14,15,23]. However, in the theoretical estimation of fluid fi lm thickness, the possible role of the sot3 tissues surrounding the joint and the influence of joint incongruity were not considered. Indeed,
it is important to emphasise the difference between the thick (order of mm), hydrostatic
fluid layer discussed here, and the thin (order of pm), hydrodynamic fluid layers predicted by theoretical analyses. It is likely that. as the thick ff uid layer is eventually depteted, a squeeze- film lubrication mechanism would develop-
The subchondral bone was considered to be rigid and impermeable, as it is much stiffer than cartilage and has a negligible time-dependent creep response. Off-axis loading and tissue cornpliance in the acetabular notch were also not included in this analysis due to the limitations of the axisyrnmetric formulation, when used in conjunction with the additional non-linearity of contact between poroelastic rnaterials. However, the observations of Terayama's extensive experiments [3 11, which were unique in their presewation of the natural hip joint anatomy during loading, lend support to the results of this cornputer analysis. Any analysis of hydrated soft tissues such as the labmm must consider the unique behaviour of such a tluid/soIid mixture. These tissues exhibit a time- dependent behaviour under applied load, and Loads are shared between the solid and fluid components of the tissue- Wu et al [35] have demonstrated that solutions obtained fiom
S.J. Ferguson Queen's University Biomechanics of the Acetabular Labrum
the ABAQUS commercial software package agree with the results of analytical investigations ofjoint contact mechanics (difFerences (8%). Others [24,32] have also shown that the results obtained fiom commercial finite element software are in good agreement (differences <4%) with the models [28] based on the biphasic theory of Mow et a[ [20], which itseIf has been validated against experimental measurements of hydrated tissues such as articular cartilage, the meniscus and the intenrertebral disc (see also
Appendix C). The finite element mesh in this study was refined until numerical convergence of the solution (changes (2%)was obtained. Any model of joint function will be sensitive to the choice of material properties for the soft tissues- Variation of the assurned cartilage properties (50% - 200% nominal) did not have an appreciable effect on the model solution; the joint response was determined by the labrurn properties. Variation of the Iabnrm material properties in this model demonstrated the sensitivity of the sealing effect of the labrurn to material properties within a range ofvalues appropriate for the tissue; a stifhess of 100 MPa for labrum tissue was required for this mechanism to occur.
The motion of the fimur into the acetabulum was determined by the redistribution of an incompressible tluid within the joint space. In the time period considered in this study, very little tluid was exchanged behveen the cartilage Iayers and the joint space.
Fluid kvas forced to the outer edge of the joint, where it pressed against the labrurn, and could return easily to its initial location once the Ioad was removed. This tluid motion and non-linear deformation of the labrurn gave the joint a certain compliance under load, and may act as a shock absorbing mechanism. As the labrum material properties change, either through ageing or disease, sofiening of the labrum through thinning of the collagen bundles [30] would lead to an increase in the compliance of the labrum. Loads applied to
SJ. Fer,won Queen's University Biomechanics of the Acetabular Labrurn
the joint would be transferred localiy to the cartilage, rather than evenly distributed across
the joint surface through fluid pressurisation.
The presence of a pressurised fluid layer behveen the cartilage surfaces coutd
prevent Wear associated with adhesion and surface shear stresses during joint motion, as
there would be no direct solid-on-solid contact over much of the joint surface. Adhesive,
shear-induced Wear due to solid-on-solid cartilage contact has been suggested as a cause
of cartilage wear [2 11. This pressurised fluid layer not only provides a low friction
articulation for the joint, but also distributes the applied load more evenIy across the
articulating surtàce. The applied load was transferred across the joint predominantly
through fluid pressure in the intra-articular gap and in the underlying cartilage, where the
unifonn distribution of interstitial fluid pressure shielded the cartilage fi-om high solid
stresses. FaiIure of this labrum seal could resuk From injury or from age-related changes
in tissue properties. Without the labrum seal, the solid matrk of the cartilage iayers must
carry a geater portion of the load. High bels of subsurface solid matrix stress may
eventually léad to tàtigue îàilure of cartilage tissue, with interna1 tàilure of the tissue
eventually propagating to the material surface causing cracks and fissures [11,3 11.
Macirowski er al [18] demonstrated, with a pressure sensing fernoral
endoprosthesis, the importance of fluid pressurisation within the acetabular cartilage
Iayer in preventing darnaging levels of cartilage solid matrix stress during Ioading. They
suggested that cartilage itself forms an effective seal against fluid expression upon direct joint contact, witli fluid forced to follow a hi& resistance path throua the hills and
valleys of the opposing cartilage surfàces. Many structures of the human body exhibit a
certain degree of redundancy, and perhaps these proposed sealing mechanisms in the hip
S J. Ferguson Queen's University Biomechanics of Uie Acetabular Labrurn
joint are not mutually exclusive. Under certain loading conditions. the analysis suggests that the labrum is able to maintain a pressurised fluid film within the joint space, providing a low fkiction articulation and shielding the cartilage layers from hi& levels of stress- If this fluid layer is depteted over time, or if such sealing would not be possible under different loadinz conditions, the labrurn may still provide a tinal seal for the high resistance fluid flow paths proposed by Macirowski et al [18].
Currently many patients with labrum damage and early arthrosis are treated with partial labral resection and joint debridement. However, it is possibk that, by compromising the sealing fimction of the Iabrum, this treatment accelerates the joint depeneration that one had hoped to alleviate. With continued study of the hnction and importance of the labrum. new surgical repair strategies can be deveIoped to maintain the overall function of the hip joint. As has been dernonstrated with improvements in meniscal defect repair, early detection and treatment of tissue damage to restore joint tùnction can delay the onset or progression of osteoarthrosis.
S.J. Fereauson Queen's University Biomechanics of the Acetabular Labrum
References
ABAQUS: ABAQUS/Standard User's ~lfanrral(version 57), Pawtucket, USA, Hibbitt, Karlsson and Sorenson, tnc., 1997a
ABAQUS: ABAQCIS/Standard neoty Mantral (version 5.7), Pawtucket, USA, Hibbitt, Karlsson and Sorenson, hc., 199%
Altenberg AR: Acetabular labnim tears: a cause of hip pain and degenerative arthritis. Sorrrhern Medical Jortrnal70: 174-1 75, 1977
Ateshian GA. Wang H: A theoretical solution for the Frictionless rolling contact of cyhdrical biphasic articular cartilage layers. Jozirnal of Biornechanics 28: t 34 1- 1355, 1995
Bullough PG, Goodfellow J, Greenwald AS, O'Connor 3: tnconconrent surfaces in the hurnan hip joint. ~'atrrre2 1 7: 1?JO, 1968
Dorrell JH, Catterali A: The tom acetabular labrum. Jonrnal of Bone and Joint Srtrgery fw68-B:400-403, 1986
Dowson D, Wright V, Longield MD: Human joint lubrication. Biomeclical Engineering 4: 160- 165, 1969
Eckstein F, von Eisenhart-Rothe R, Landgaf J, Adam C. L6he F, MUller-Gerbl M, Putz R: Quantitative analysis of incongruity, contact areas and cartilage thickness in the human hip joint. rlcrcr Anuromica 158: 192-204, 1997
Fein RS: Are synovial joints squeeze-film hbncated? Proceedings of the lnsrirzrtion of :Mecltanicol Engineers. Jorrrnal of Engineering in me ni ci ne [Hl 18 1: 1 25- 12 8. 1967
Fithian DC, Kelly MA, Mow VC: Material properties and structure-fùnction relationships in the menisci, Ciinical Orrhopaedics and Related Resear-ch 253: 1 9-3 1, 1 990
Freeman MAR: 1s collagen fatigue failure a cause of osteoarthrosis and prosthetic coinponent migration? A hypothesis. Jorrrnd of Orthopaedic Research 1 7:3-8, 1999
Harris WH, Boume RB, Oh 1: Intra-articular acetabular labrum: a possible etiologïcal tàctor in certain cases of osteoarthritis ofthe hip, Joztrnaf of Bone undJoint Sur-geq 6 1- A510-5 14, 1979
Helmig P, Sojbjerg JO, Sneppen O, Loehr JF, Ostgaard SE, Suder P: GIenohumeral movernent patterns afier puncture of the joint capsule: an experimental smdy. Journal of Shordder and Elborv Srrrgeiy 2209-2 15, 1993
Higginson GR: Elastodynarnic lubrication in human joints. Engineering in Medicine 7:35- 41, 1978
Higginson GR, Norman R: The lubrication ofporous elastic solids with reference to the functioning of hurnan joints. Jozrrnal of :VfechanicalEngineering Science 16:250-257, 1974
S.J. Ferguson Queen's University Biomechanics of the Acetabular Labnim
Ikeda T, Awaya G, Suzuki S, Okada Y, Tada H: Tom acetabular labrurn in young patients. Arthroscopie diagnosis and treatment. Jorrrnal of Bone and Joint Srwgery [Br- 70-B: 13 - 16, 1958
Kurnar VP, Balasubramaniam P: The role of atmospheric pressure in stabilising the shoulder. Journal of Bone andJoinr Srrrgery [Btj 67-B:7 19-72 1, 1985
Macirowski T, Tepic S, Mann RW: Cartilage stresses in the human hip joint, Jorrrnnl of Biornechmical Engineering 1 16: 10-1 8, 1994
McCarthy JC, Busconi B: The role of hip arthroscopy in the diagnosis and treatment of hip disease. Canadian Jorrrnal ofSrrrgety 38 Suppl 1:S13-17, 1995
Mow VC, Kuei SC, Lai WM, Armstrong CG: Biphasic creep and stress relaxation of articular cartilage in compression: theory and experiments. Jotrrnal of Biomeclranicd Engineering 1 0273-84, 1980
Mow VC, Soslowsky LI: Lubrication and wear of joints. In: Basic orhpaedic biontechanics, 245-292. Ed by VC Mow and WC Hayes. New York, Raverr Press, 199 1
NLM. The Visible Human Project. National Library of Medicine . 1999. (http://~~~~~.nlm.~ih~gov/research/visib1e~humm~html)
OfKellyJ, Unsworth A, Dowson D, Hall DA, Wright V: A study of the role of synovial tluid and its constituents in the friction and lubrication of human hip joints. Engineering in Medicine 7:73-83, 1978
Prendergast PJ, van Driel WD, Kuiper JH: A cornparison of finite elernent codes for the solution of biphasic poroelastic problems. froceedings of rhe Itwirrrrion of 12.iecl~nnicul Engineers. Jo~rrnalof' Engineering in Medicine [Hl 2 10: 13 1- 136, 1996
Proctor CS, Schmidt MB, Whipple RR, Kelly MA, .Mow VC: Material properties ofthe normal medial bovine meniscus, Journal of Orrhopnedic Research 7:77 1-78?, 1989
Shibutani N: Three-dimensional architecture of the acetabuiar labrurn - a scanning electron microscopic study. Jorrrnal of ille Japanese Orthopaedic Association 6232 1-329, 1988
Soltz MA, Ateshian GA: Experùnental verification and theoretical prediction of cartilage interstitial fluid pressurization at an impermeable contact interface in confined compression. Jorrrnal of Biomechanics 3 1 :937-934, 1998
Spilker RL, Suh JK, Mow VC: A finite element formulation of the nonlinear biphasic mode1 for articular cartilase and hydrated sol? tissues including strain-dependent permeability. In: Cornplriarional hfethocisin Bioengineering., 8 1-92. Ed by RL Spilker and BR Simon. New York, American Society of Mechanical Engineers, 1988
Takechi H, Nagashima H, Ito S: Intra-articular pressure of the hip joint outside and inside the limbus. Jorrrnal ofthe Jcrpanese Orthopaedic Association 56529-536, 1 982
Tanabe H: Aging process of the acetabular iabnun - an electronic microscopic study. Jorrrnul of rire Japanese Orrhopaedic Association 65:18-25, 199 1
S.J. Ferguson Queen's University Biomechanics of the Acetabular Labrum
Terayama K, Takei T, Nakada K: Joint space of the human knee and hip joint under a static load. Engineering in Medicine 9:67-74, 1 980
van der Voet AF, Shrïve NG, Schachar NS: Numerical mode1hg of articula- cartilage in synovial joints - poroelasticity and boundary conditions. In: Comprrter Merirods in Biomechanics and Biomedical Engineering, 200-209,Ed by J Middleton, GN Pande, and KR Williams. Swansea, IBJ Publishers, 1993
Warner JJP, Deng X, Warren RF, Torzilli PA, O'Brien SJ: Superoinferior translation in the intact and vented glenohumeral joint. Journal of Shorrider and Elbow Srwgery 299-205, 1993
Weber W, Weber E: Uber die mechanik der menschlichten gehwerkzeuge nebst der beschreibung eines versuches uber das herausfallen des schenkelkoptès aus der pfanne im luftverdennten raum. Annalen Physik rlnd Cilemie 40: 1-1 3, 1837
Wu JZ, Herzog W, Epstein M: Evaluation of the finite element software ABAQUS for biomechanical modelling of biphasic tissues. Jorrrnal of Biomechanics 3 1 :165-1 69, 1998
S J . Ferguson Queen's University Chapter Six
The Influence of the Acetabular Labrum on
Hip Joint Cartilage Consolidation:
a Poroelastic Finite Element Model
S.I. ergu us on'-', J.T. ~r~ant-',R. ~anz' and K. tto'.'
' AO AS IF Research Institute, Davos Platz, Switzerland
'~e~artmrntof Orthopardic Surgery, University of Berne, Inselspital, Berne, Switzerland
'~c~artrnento F Mechanical Engineering, Queen's University, Kingston, Canada
In Press: Journal of Biornechanics (2000) Biomechanics of the Acetabular Labrum
Abstract
The _goal ofthis study was to investigate the influence of the acetabular labrum on the consolidation, and hence the solid matrix strains and stresses, of the cartilage layers of the hip joint. A plane strain finite element model was developed, which represented a coronal slice through the acetabular and femoral cartilage layers and the acetabular labmm. Elements with poroelastic properties were used to account for the biphasic solid
/ fluid nature of the cartilage and labrum- The response of the joint over an extended period of loading ( 10,000 seconds) was examined to simulate the nominal compressive load that the joint is subjected to throughout the day. The model demonstrated that the labrum ad& an important resistance in the flow path of fluid being expressed fiom the cartilage layers of the joint. Cartilage layer consolidation was up to 40% quicker in the absence of the labrum. Following rernoval of the labrum £kom the model, the solid-on- solid contact stresses between the femoral and acetabular cartilage layers were greatly increased (up to 92% higher), which would increase fiction benveen the joint surfaces. in the absence of the labrum, the centre of contact shifted towards the acetabular rim, overloading cartilage that may be previously conditioned to Iower levels of stress.
Subsurface strains and stresses were much higher without the labrum, which could contribute to fatigue damage of the cartilage layers. Finally, the labmm provided some structural resistance to lateral motion of the femoral head within the acetabulum, enhancing joint stability.
Keywords: hip joint, labrum, finite element, poroeIastic, cartilage
SJ. Feraouson Queen's University Biornechanics of the Acetabular Labrunt
Introduction
The causes of osteoarthrosis (OA) in the hip are not fully understood, but it has been proposed that contributory factors are changes in the mechanical environment of the joint and changes to the mechanical properties of the articular cartilage [32,33,38,41].
The acetabular labrum is a fibrocartilaginous lip with a tissue structure similar to the rneniscus of the knee. The majority of the labrurn is composed of thick, Type I collagen
fibre bundles principally arranged parallel to the acetabular rim, with some fibre bundles scattered throughout this layer running obliquely to the predominant fibre orientation
[48]. It is attached to the osseus margin of the acetabulum, deepening the acetabular socket and extending the covenge of the femoral head (Figure 6.1). As any discussion of the nature and oripin of OA should consider, among other factors, the function of the accessory structures of the joint [26], one may speculate that an abnormal acetabular labnirn may be implicated in the aetioIogy of joint degeneration. Several studies have shown that labrum excision, or pathology of the intact labrum, such as tears or intra- articular impingement, may lead to joint changes consistent with ear1y OA
[S,2 l.27,28,34]. Cartilage fibrillation has been observed directly adjacent to labrum defects in post-rnortem hip joint specimens [ L 8,441, Ikeda et al [29] postulated that tears in the labrum compromised the load bearing and stability enhancing function of the labrum. Before making predictions about the consequences of labrum pathology, however, it is important to understand the role of the acetabular labrum in a normal, healthy joint.
S.J. Fer,o;uson Queen's University Biomechanics of the Acetabular Labrum
Figure 6.1: The acetabular labrum ( 1) is a tÏbrocarti1aginous rim attached to the margin of the acetabular socket, with a tissue structure similar to the meniscus. The labnim joins smoothly with the lunate surfiace of the articuiar cartilage (2), deepenine the acetabutar socket. lt joins with the transverse acetabular ligament (3). bridging the acetabular notch and forming a complete circle. It is triangular in cross-section, with its base attached to the acetabulum and its apex fonning the tkee edge of the labrum, which is turned in against the fernorai head.
The mechanical role of the labrum in normal hip function is not well understood.
Several experimental studies have shown that the labrum can provide a scal against fluid flow in and out of the intra-articular space [55,56,59]. Besides improving the stability of the joint through a vacuum effect, this possible sealing function of the labrum could enhance lubrication mechanisms in the hip joint. The ability of the labrum to contain a pressurised fluid layer within the hip joint under simple loading conditions was the focus of a previous study [23j. This analysis dernonstrated that the labrum could seal a fluid
S.J. Ferguson Queen's University Biomechanics of the Acetabular Labrurn
layer within the joint for a period of several minutes. This pressurised fluid layer
prevented direct contact of the joint surfaces and distributed the applied load more evenly
across the cartilage surf'aces.
The labrum may play another role in normal hip joint function independent of its
ability to seal the intra-articular space of the joint. The labrum may increase the load-
bearing surface of the hip joint, decreasing the overall stress level in the cartilage. With a
relatively low permeability, the labrurn may limit the rate of fluid expression from the
cartilage layers dunng loading. McCutchen [36] demonstrated, in an analytical mode1 of joint contact, that applied loads are partitioned within cartilage, due to its mixed solid /
fluid composition. Cartilage is able to carry loads through fluid pressurisation for
estrernely long periods of time, because of the high resistance to Aow of interstitial fluid
out of the tissue. With the majority of load carried by fluid pressure, the magnitude of
stresses within the collagenous solid matrix of the cartilage is limited [7,50,5 11. Fluid
pressurisation in cartilage is not limited to short duration loading; interstitial fluid
pressures in congruent joints could theoretically persist for hours. Using finite element
models, others have demonstrated the importance of this tluid pressurisation mechanism
in cartilage [20,30,57], Fluid pressurisation lirnits the magnitude of contact solid-on-solid
stresses (ie. the stress behveen contacting asperities of the mesh-like collagenous solid
matrix), Iessening friction at the cartilage surfaces [30,35,58]. Failure of this mechanism
could Iead to increased fiction and higher loading in the solid matrix of the cartilage
surfaces.
Our hypothesis is that the labrum adds an important resistance in the tlow path of
fluid expressed from the cartilage layers, enhancing this protective mechanism for the
S.J. Ferguson Queen's University Biomechanics of the Acetabular Labrurn
------tissue. The goal of this study was to develop a computer model to study the influence of the labmm on hip joint cartilage deformation, fluid pressures and solid stresses during extended loading periods. In the absence of the acetabular labrurn, do the cartilage layers of the hip joint consolidate more quickly, and does this result in higher levels of local cartilage deformation and stress?
Method
For this study, the commercial finite element sofcware ABAQUSwas chosen for its ability to simultaneously model contact mechanics and poroelastic (Le. fluid saturated porous elastic) materials [1,2]. A two-dimensional, plane-strain finite element mode1 was developed. The model represents a coronal slice through the hip joint (Figure 6.2) to allow loads approximating the average physiologicat loading direction. Representative hip joint geometry was based on average measurernents from several clinical MRI data sets of non-symptomatic hip joints (Department of Orthopaedic Surgery. lnselspital
Berne), The femoral head had a radius of 26rnm, the opposüig cartilase layers were 3mm thick. and the underlying bone kvas considered rigid and impermeable. Dalstra er al [19] have shown. with an experimentally validated finite element mode1 of the entire pelvis. that the stresses in the cortical shell of the acetabulum, for an applied load of approximately 0.75 - 1 times bodyweight, are several orders of magnitude less (0.04%) than the modulus of the cortical bone itself. Therefore, one can infer from this that the deformation of the bony structures would be negligible.
S.J. Ferawon Queen's University Biomechanics of the Acetabular Labmm
Figure 6.2: Plane-strain finite element mesh of a portion of the acetabular (1) and fernoral(2) cartilage surfaces and labrum (3). The circumferential (out-of-plane) stiflhess of the labrum is simulated with in-plane tws elements aligned with the direction of the acetabular rim. The acetabulurn is fised and a vertical load is applied to the fernur. Output variable sampling points (A,B,C) are indicated (see Results).
S.J. Ferguson Queen's University Biomechanics of the Acetabular Labrum
The cartilage was modelled as an isotropic poroelastic material. The Lamé
constants of the solid phase, 1 and p, as defined by biphasic mode1 of Mow ef [37]
were converted to the poroelastic matenal properties E and v by:
The permeability k, as defined in the biphasic model [37], were converted to that of the
ABAQUS poroelastic model k' by:
where y is the specific weight of the pore fiuid. The properties assumed for cartilage
were consistent with previous analyses ofjoint contact mechanics [57,6 11: moduIus E =
0.467 MPa, permeability k = 7.358x10'~mm/s,Poisson's ratio of the drained solid ma&
v = 0.167, specific weight of interstitial fluid y= 9.8 lxlo4 ~/rnm'and solid fraction $s =
20 percent of the total tissue volume. The labrum was modelled as an isotropic
poroelastic material with properties (E = 0.2 MPa, k = 2.0~1 0-8 mrn/s) similar to those of
the meniscus [24,43], and to those observed in prelirninary confined compression tests of
labrum tissue conducted in Our lab. As an initial condition for this model, there was no
interposed fluid layer benveen the joint surfaces, with direct contact between the two
cartilage layers. Contact elements were inserted between the articuIar surfaces.
Fluid expression was permitted From any exposed (non-contacting) surfaces of the cartilage layers and Iabrum. Cartilage surfaces which corne into contact would be regions
S.J . Ferguson Queen's University Biomechanics of the Acetabular Labnun
of very low permeabilicy to fluid flow [30,60];therefore, contact regions were sealed against fluid expression normal to the surface. These boundary conditions were updated throughout the solution to account for changes in contact area.
To simulate the out-of-plane circumferential stifhess of the Iabrum in this two- dimensional model, tmss elements joined the labrum elements to the geometric centre of the acetabulum. The stifiess of these tmss elements kvas varied over a range from 5 -
100 MPa, with a basetine value of 20 MPa. This baseline stiffhess was chosen to give the plane-strain labrum model the same initial bending deflection per unit load as that calculated using a simple, linear-elastic, three-dimensional model of the labrum ring.
The stiffness and permeability of the labrum were vaned pararnetrkally (50% - 200% of nominal values) to test the scnsitivity of the model to uncertainty in these properties. A vertical load was applied to the femoral component of the model and then maintained.
This load produced an initial contact pressure in the acetabular cartilage equal to that calculated in an axisymmetric poroelastic joint contact model subjected to a load ~f 0.75 tirnes bodyweight. The model results were collected over the time period from O - 10,000 seconds to evaluate the rate of cartilage layer consolidation, stresses in the cartilage solid matrix, and tluid pressure in the cartilage layers for models with and without the labrum.
S.J. Ferguson Queen's University Biomechanics of the Acetabular Labrum
Resutts
The results from this finite element analysis demonstrated the influence of the
Iabrum on hip joint cartilage consolidation over an extended loading perïod. While the stiffness and penneability of the labmm tissue was varied pararnetrically, the changes which resulted From these variations were srna11 (~3%difference) compared to the differences observed in the presence or absence of the labrum. Therefore, only the results of the model with a labrum and with baseline material properties were compared to those of the mode1 without a labrum,
Afier imposition of the load, the femur moved superiorly within the acetabulum.
As the acetabulum opens laterally, but is reinforced medially by the pelvic bones, the femur also shifted laterally under load. These displacements increased with time (creep) as interstitiat fluid was forced fiom the joint surfaces (Figure 63). The creep- consolidation rate of the hip joint cartilage layers was calculated as the rate at which the femur and acetabulum approached each other (i.e. the tlrst derivative of the displacement- time curve of the centre of the fernord rnesh rigid bone elements relative to the centre of the acetabular mesh rigid bone elernents). This was up to 40% faster in the joint without a labrum, resulting in pater cartilage compression over time. After 10.000 sec the cartilage Iayers had compressed 35% more in the model without a Iabrum, and the femur displaced further laterally relative to the mode1 with an intact Iabrum. These time dependent differences in the creep-consolidation of the cartilage layers were reflected in the derived quantities of total contact pressure, solid-on-solid contact stress, and strain and stresses within the solid matrix.
S J. Ferguson Queen's University Biomechanics of the Acetabular Labrum
Vertical Joint Consolidation
O 2000 4000 6000 8000 10000 tirne (sec)
Figure 6.3a: Cartilage layer consolidation. Vertical motion of the femur relative to the acetabulum has been pIotted for rnodek with and without an acetabular labrum. The cartilage layers cornpress approximately 40% quicker if the labrum is removed.
Lateral Joint Consolidation
4 labrum
-IlIII O 2000 4000 6000 8000 10000 time (sec)
Figure 6.3b: Cartilage layer consolidation. LateraI motion of the femur relative to the acetabulurn has been plotted for rnodels with and without an acetabular labmrn. The labnim limits motiori of the krnur within the acetabulum.
S-J. Ferguson Queen's University Biomechanics of the Acetabular Labrurn
The distribution of total contact pressure alonç the acetabular surface at t = 1,000 seconds is shown for the joint with and without a labrum (Figure 6.4). The total contact pressure at a point is the sum of the solid-on-solid contact stress and the interstitial fluid pressure at the interface between the femur and the acetabulurn. There was a parabolic distribution of the total contact pressure in the acetabular roof, with a higher peak pressure in the model without the labrum. The mavimum total conract pressure, interstitial fluid pressure and solid-on-solid contact stress calculated dong the acetabular surface are summarised in Table 6.1, which also shows how these values change with time. The maximum contact stress, total contact pressure and fluid pressure do not always occur in the same location for a given mode1 and time step, due to the uneven consolidation of the cartilage Iayers. Therefore, the peak values given in the table do not surn directly, as the stresses and pressures at a single point would. The ratio of interstitial fluid pressure to solid-on-solid contact stress was calculated in the highly loaded zenith of the acetabulum. This provided an indication of how the load was partitioned between the solid and fluid constituents of the cartilage tissue. At t = 1,000 seconds, approximately 94% of the load at that location was transferred through fluid pressure, and not through direct solid-on-sotid contact stresses, for the model with the intact labrum, while 92% of the load was transferred through fluid pressure for the model without the labrum. After 10,000 seconds, the fraction of the load carried by the fluid phase of the tissue was 80% and 73% for the mode1 with and without the labrum, respectively.
S.J. Ferguson Queen's University Biomechanics of the Acetabular Labrum
Figure 6.4: Total contact pressure (interstitial fluid pressure plus direct solid-on-solid matrix contact stresses) exened by the femoral cartilage layer on the acetabular cartilage Iayer at t = 1,000 seconds aller load application, Total contact pressure with the intact labmm is plotted in dark gey; total contact pressure without the Iabrurn is plotted in Iight grey as a semi- transparent overlay. Position reference marks are plotted in five degree intervals from the osseus acetabular rim. A 200 kPa scale bar is provided for reference, There was a higher total contact pressure in the absence of the labrurn,
t = 1,000 sec t = 10,000 sec
With Without With Without Labrum Labrum Labrum Labrum
Soiid Contact Stress os(Wa) 30.0 49.3 77.8 149-2
Fluid Pressure P (kPa) 351.9 404.6 276.4 397.2
Total Contact Pressure (kPa) 3 73.1 438.5 339.7 40 1.3
Load carried by Fluid (%) 94.1 92.2 80 -0 73 .O
Table 6.1 : Peak solid-on-solid contact stresses, interstitial fluid pressure, total contact pressure (kPa) and load partitioning at the articular surface- The locations of local stress and pressure maxima do not necessarily coincide. The percentage of the load camed by fluid pressure provides a measure of how the applied load is partitioned between the solid matrix and fluid phase of the tissue.
S.J. Fer,ouson Queen's University Biomechanics of the Acetabular Labrum
The labrum infiuenced not only the short-term response of the joint to load, but also the gradua1 creep behaviour of the cartilage layers. The solid-on-solid contact stresses behveen the femur and acetabulum increased over time as fluid was expressed from the cartilage layers and the tissue consolidated (Figure 6.5). These stresses have been plotted along the acetabular surface for the joint with and without the labrum, at t =
1.000 and t = 10,000 seconds afler imposition of the step load. The solid-on-solid contact stresses were up to 64% higher at t = 1 ,000 seconds and 92% higher at t = 10,000 seconds in the absence of the labrum, initial contact was between the labrum and kmur, at the edge of the joint. The proxirnity of the fiee-draining outer surface of the labrum minimised the effectiveness of load carriage through fluid pressurisation, as there is little resistance, due to the short path length, to tluid expression frorn the labrum. Therefore, the solid-on-solid contact stresses in the labrum are initially hi&, due to the fast consolidation of the tissue in this area.
The values of the total contact pressure (tluid plus solid) along the acetabular cartilage surface. from the fossa to the cartilage / labrum interface, have been plotted in
Figure 6.6. Not only was the peak contact pressure approximately 18% higher for the joint without a Iabrum, but the centre of pressure shifted laterally when the labrum was removed, up to 13O at t = 10,000 seconds after load application, relative to the centre of pressure in the intact joint. This tendency was also observed in the solid-on-solid contact stresses. This kure shows the maximum and minimum contact pressures which were calculated for the intact joint during parametric variation of the stifThess and pemeability
(50 - 200% of nominal values).
S J, Fer,ouson Queen's University Biomechanics of the Acetabular Labrum
+ Joint clilw
- Joint angr
Figure 6.5: Solid-on-solid cartilage contact stress has been plotted (a) at t = 1.000 seconds and (b) at t = 10,000 seconds after load application. Solid-on-solid contact stress for the mode1 with the intact labnirn is pIotted in dark grey; solid-on-solid contact stress for the mode1 without the Iabmm is plotted in Iight grey as a semi-transparent overlay. A 100 kPa scale bar is provided for reference. The solid-on-solid contact stresses in the acetabular cartilage increase with rime, and are up to 92% higher in the absence of the labrum.
S.J. Ferguson Queen's University Biomechanics of the Acetabular Labmm
Total Contact Pressure
Position on Acetabulum Relative to the Osseus Rirn (degrees)
Figure 6.6: Total contact pressure (interstitial fluid pressure plus solid-on-solid contact stress) at t = 10,000 seconds after load application. The maximum (dashed line) and minimum (dotted line) total contact pressures calculated during parametric variation of the Iabrum rnaterial properties (50% - 200% of nominal values) are plotted for cornparison to the total contact pressure calculated for the model wvithout the Iabrum. The effect of labrurn removal is much greater than the effect of rnaterial property uncertainty. The peak contact pressure was 1 8% higher witliout the Iabrum, and the centre of pressure shifted 13 degrees laterally relative to the centre of pressure in the intact joint (see Figures 6.4 and 6.5 for angular reference marks)-
The differences in the model results due to uncertainty in the material properties were small compared to those observed when the Iabrum was removed fiom the model.
The influence of the labrum on stresses and fluid pressures was not confined to the interface between the two articular surfaces. Similar changes were observed in the cartilage deformation and the resulting stress state beneath the joint surfaces.
Representative stress and strain profiles, with and without the labrum, are presented in
Figures 6.7 - 6.10.
S.J. Fer,ouson Queen's University Biomechanics of the Acetabular Labm
(Ave. CnLr 75%) +2.6&&3-01 &300e-af r2.07W-01 -1 .B40e-01 +1.610801 tl.38060l tl.lSW-01 -9200eM t6.900e-02 cO.ôOQeû2 +2300e-02 +O.MX)e-OO
E Max. Principal (Ave. Cnt; 75%') 4.738401 +2WOe-01 +1.800e-01 +1.600e-01 +l.4Ooe-Ol +1 a0401 +1.Dooe-Oi +8.000e-02 +6.Wk-û2 4.000e-02 +2cmce-a2 +0.0aae+Oo
Figure 6.7: Maximum tensile principal strain in the cartilage Iayers with (a) and withoot (b) the labnim, IO00 seconds after load appIication. Local strain levels at the bone cartilagehterfaces are pater in the absence of the labnim.
S.J. Fer-mon Queen's University Biomechanics of the Acetabuiar Labrum
(Ave Cni :75%) 43511&1 ROM3eOl +18o08-01 +1.600%-01 -1.400eQI +1.200e-01 cl.uooe-O1 -6.ûûûû-02 6.ûûûû-M 4.Ooe-M -2.000602 -3.068e-08
Figure 6.8: VonMises equivalent stress IeveIs with (a) and without (b) the labrum, 1000 seconds afier Ioad application. Areas of high stress at the cartilage/bone and labrumhone interfaces are more extensive in the absence of the labrum. Units are MPa.
S.J. Ferguson Queen's University Biomechanics ofthe Acetabular Labrum
S. Tresca (Am Cnt: 75%) 4.750e-01 4.000c-01 +2700P01 -2MOe-01 +2100e01 -1.BM)e-Ol -1 SOOe-01 -1.200e-01 -9.000e-02 -6.aaOe-02 4.WOe-M -2659e-QB
S. Tresca 1 (Aue. Cm.: 75%) ci.537e-01 +3.000e-01 r2700e-01 r2400e-01 r21OOB-Ol cl .BOOe-01 +t .MCe-ût *12oh-o1 +9.000e-02 4.000402 +3.O00e-<)2 4.195606
Figure 6.9: Tresca "stress"levels with (a) and without (b) the labium, 1000 seconds after load application. The Tresca stress is defined as the maximum difference behveen principal stresses, and so is double the maximum shearing stress, Units are MPa.
S .J . Ferguson Queen's University Biomechanics of the AcetabuIar Labrum
Figure 6.1 0: Compressive principal stresses with (a) and without (b) the labrurn, 1000 seconds afrter load application. Units are MPa.
S.J. Ferguson Queen's University Biomechanics of the AcetabuIar Labm
A cornparison of representative mode1 results with and without the labrum, for points throughout the cartilage layers, is summarïsed in TabIe 6.2. Local stresses and strains were consistently higher (25 - 4 1 %) in the absence of the tabmm, at t = 1,000 seconds,
With brithout With Without With Without Labrum Labrum Labrum Labmm Labrum Labmm -. Maximum Compressive -0.130 -0.163 -0,062 -0.084 -0.3 1 1 -0,426 Principal Stnin (el)
Maximum TensiIe 0.1 15 O. 144 0.034 0.048 0.229 0.3 19 Principal Strain (E;)
Von Mises Equivalent 85 I 07 34 46 188 259 Stress (kPa)
Maximum S hearing 49 62 20 37 108 149 Stress (kPa)
Table 6.2: Sub-surface strain and deformational stressesi in the cartilage layen at t = 1.000 sec. Values were calculated at points (A) and (B) Imrn below the articuIar surface, in the highly Ioaded superior acetabulum region and (C) at the cartilage / bone / Iabmm interface. Compressive principaI strains are negative.
' Von Mises "equivaienr stress" is deiïned as:
îi-om the Von Mises yield criterion for ductile matenals. which predicts that a material niII yield when the calculated Von Mises equivalcnt stress at a point reaches the yield stress of the material in uniaxial tension. The yicld criterion itself is based on the distortional energy within a body at the point of yielding.
S.J. Fermouson Queen's University Biomechanics of the Acetabular Labrum
The velocity of thid flowing tangentially. through the cartilage layers, \vas an
order of magnitude greater than that of fluid flowing across the layers (ie. perpendicular
to the articular surface). The predominant direction of fluid flow, therefore, was paraIIel
to the cartilage surface and across the labrum, ttom the centre of contact towards the
perirneter of the joint. The maximum fluid velocity \vas higher (6.4 x IO-' mdsvs. 3.27
x IO-' mm/s) in the absence of the labrum.
Discussion
The mode1 predicted that the presence of the labrum lirnited the level of cartilage
de formation and stress through a sealing mechanism. The labrum added an additional
resistance in the flow path of fluid being expressed ftom the cartilage iayers of the hip
joint, enhancing the retention of interstitial fluid within the tissue. Since cartilage layers
detom predominantly through changes in the tissue volume, and this occurs through
t'luid expression due to the assumed incompressible nature of its solid and tluid
constituents CLIO], any additional resistance to fluid flow from the cartilage layers slows
the rate of defornation. Therefore, the labnrm enhances the ability of the cartilage layers
to carry loads by interstitial fluid pressurisation, which limits the stresses within the
collagenous solid rnatrk [5 11. The importance of such a sealing rnechanism, whereby the
majority of the load applied to a joint is transferred across the cartilage layers through
fluid pressure rather than through solid matrix stress, has been demonstrated in a variety of theoretical joint contact models [20,30,45,52,57]and also by direct experimental measurement [511.
S.J. Ferguson Queen's University Biomechanics of the AcetabuIar Labnrm
Numerical analyses of hydrated tissues, based on the biphasic model of Mow et al
[37], have demonstrated the importance of considering the unique solid / tluid
composition of such tissues. Mode1 solutions obtained from commercial finite element
software [42,57,6 11 have show good agreement with models based on the biphasic
theory [53]. Indeed, the poroelastic theory used in commercial finite element software
for the analysis of wet soils is equivalent to biphasic theory if the efiècts of fluid viscosity
are neglecred [16,17,49]. The model used in this study was only an approximate
representation of the cornplex structure of the human hip joint. A plane-strain model was
developed to allow load application in a physiological direction. An axisymmenic model
with non-axisymmetric loading or a full three-dimensional model would provide a better
representation of the hip joint geornetry. However, such a model. with the addition of
poroelastic cartilage layers and contact prediction, would be extremely cornplex.
Suficient material exists on either side of the plane of the model to limit out-of-plane
strains, and interstitial tluid tlow follows the direction of the maximum pressure gradient,
which is in the plane of the model. Although simplified, this rnodel provided a prediction
of the influence of the acetabular labrum on hip joint cartilage consolidation. The consolidation behaviour of the cartilage layers was exarnined over a long loading period, both with and without an intact labrum.
The loading for this analysis was based on the in vivo hip force measurements taken by Berpann er al for a vanety of daily activities [I 1-15]. These measurements showed that the joint is subjected to a variety of short-duration, high magnitude loading throughout the day. These measurements also demonstrated that there is always a residual compressive force acting across the hip joint, even during sedentary periods, and that on
S.J. Ferguson Queen's University Biomechanics of the Acetabular Labm average the compressive force across the joint is equal to or greater than bodyweight, and directed within a narrow angular range about vertical. The poroelastic nature of the cartilage layers dictates that under high magnitude dynamic or impact loading, forces are transferred across the joint almost exclusively by pressure within the interstitial fiuid. A short-duration Ioad alone could not cause sufficient expression of fluid from the cartilage
Iayers, and tissue deformation, due to the long time constant for cartilage consolidation, which is determined by the rate of fluid expression. However, the action of a constant, low-magnitude compressive force would be sufficient to cause substantial tissue consolidation, with consequent increases in the strain and stress levels within the coIIagen matrk of the tissue. Therefore, this study was focused on the response of the joint under the influence of a long-term, constant Ioading. Tt is possible that there is only a limited opportunity for fluid, driven by a relatively weak 0.2 MPa osmotic pressure, to return to cartilage layers during daily activities. While the cartilage layers are well adapted to the dsmands of short duration loading, a protective mechanism may be necessary to aliow the cartilage layers of the joint to retain their interstitial fluid during extended periods of loading.
The efficient sealing of fluid paths from the cartilage layers can be inferred from the velocity of interstitial Auid Aow. The fluid velocities predicted by the mode1 were of the same order of magnitude ( IO-' rnm/s) as those calculated from in sihr femoropatellar cartilage volume change measurements by Eckstein el al [Z 1 b], and by the simulations of confined compression testing in Appendix C. It is unlikely rhat such £luid flows of less than 1 pdscould mechanically damage the cellular structure of the cartilage layers or lead to proteoglycan mig-ation, when one considers the relative size of the the celluIar
S.J. Fermouson Queen's University Biomechanics of the AcetabuIar Labrum
and collagenous components of the tissue, for example (see Figure 2.8) and the viscosity of the fluid. However, the interstitial fluid flows related to low fiequency loading have been proposed as stimulatory factors in proteoglycan aggrecan synthesis [17b]. In vitro expenments have demonstrated that fluid velocities greater than 1pm/s may enhance the transport of inhibitors of matrix-degrading enzymes [36bj. While fluid velocities of the same order of magnitude as those predicted in the mode1 have been cited in such studies, the relationships between tluid velocity and the cellular processes are not welI understood. As the boundaries between positive and negative stimuli are not yet clear, it would be difficult to speculate whether the differences in fluid velocity predicted by the mode1 with and without the Iabrum are significant.
Nevenheless, the maintenance of fluid pressurisation within the cartilage layers of the joint has important consequences in the normal îûnction of the joint. With the majority of the load being supported by fluid pressures, the actual solid-on-solid contact stresses are low, which may account for the extremely low friction characteristics of the joint [XI. With the labrum removed, the solid contact stresses behveen the cartilage
Iayers were up to 90% higher followingjouit consolidation. This higher Ievel of soIid contact stress wodd iead to increased Enction during joint motion, which has been suggested as a cause of cartilage wear 1391, or even heat-induced changes to the celluIar structure of the cartilage itself [3 II. The high levels of cartilage deformation predicted by the mode1 have been observed in sihr in measurements of cartilage layer deformation in response to prolonged loading [2 1b]. Below the articular surface, shear stresses due to cartilage deformation were up to 27 - 38% higher throughout the cartilage layers following labrum removal. This elevated ievel of tissue deformation could contribute to
S.J. Fer,ouson Queen's University Biomechanics of the Acetabular Labrum fatigue failure of the coliagen fibre mamx of the tissue, and eventually to cartilage fibriltation, crevassing or delamination [25,39]. indeed, the elevated levels of shear stress predicted at the cartilage / labt-um / bone interface correspond to the location of cartilage delamination observed in post-mortem examinations [18]. Once shear failure has occurred at this location, the cartilage-bone interface would no longer be impervious.
The cartilage / bone interface would become free drainin= and a new pathway would be available for interstitial fluid flow, thus reducing the effectiveness of the mechanism of
Ioad support through fluid pressuristion. The applied load would be transferred to the solid matrk in the region of this defect, leading to further damage [6].
Not only were the magnitudes of the calculated tissue strains and stresses higher in the absence of the labrum, but their distribution within the joint was altered as well.
With an intact labrum, the total contact pressure and the direct solid-on-solid contact stress were more evenly distributed across the joint surface than with the Iabrum removed. FolIowing Iabrum removal, the centre of contact pressure shit'ted towards the acetabular rim. This shift in the contact pressure distribution could be a contributory factor in cartilage degeneration. The mechanical properties of the cartilage in the hip joint
Vary with location [8,9]. It has been proposed that there is a spatial adaptation of the cartilage in diarthodial joints to its mechanical environment. that the cartilage from highly loaded areas of the joint is generally stiffet than that frorn areas which do not regularly esperience hi& loads [54]. The shifr of the contact pressure applies a greater than normal load to the cartilage in the lateral portion of the joint. Overloading of cartilage, which has been pre-conditioned to lower levels of stress, has been suggested as a contributing factor in osteoarthrosis of the patello-femoral joint [47].
SJ. Ferguson Queen's University Biomechanics of the Acetabular Labrum
The predictions of this study imply that the labrum contributes to overall joint
function, by limiting the rate of cartiIage Iayer consolidation, thus reducing the solid-on- solid contact stresses behveen opposing cartilage surfaces. Current treatment methods for
labrurn pathology, e-g. partial labral and joint debridement, rnay compromise this sealing
function of the labrum. It is possible that this treatment accelerates the joint degeneration that one had hoped to alleviate. As has been demonstrated in the case of meniscal tears, total resection of the darnaged tissue can lead to the development of premature OA
[4,22], due to a loss of the the specific function of the meniscus in load distribution within the joint [3,10,46]. However, proper repair of tissue damage can hl 1y restore joint function [ 101. With continued study oE the îùnction and importance of the labrum, new surgical repair strategies may be deveIoped to maintain the overaIl fùnction of the hip joint.
S.J. Ferguson Queen's University Biomechanics of the Acetabular Labrum
References
AB AQUS: ABAQUS/Standard User5 Manzrcrl (version 5.7). Pawtucket. USA. Hibbitt, Karlsson and Sorenson, Inc., 1997a
ABAQUS: ABAQUS/Srandard T?zeory ~tlanzral(version 5.7), Pawtucket, USA, Hibbitt, Karlsson and Sorenson, hc., 1997b
Ahmed AM, Burke DL: in-vitro measurement of static pressure distribution in synovial joints - Part 1. Tibia1 surface of the knee. Jozrrnai of Biomechmical Engineering 105:216- 225, 1983
Allen PR, Denham RA, Swan AV: Late degenerative changes after meniscectorny. Factors affecting the knee afier operation. Jozrrnal of Bone andJoint Szrrgery [Br] 66-B:666-67 1, 1984
Altenberg AR: AcetabuIar labrum tears: a cause ofhip pain and degenerative arthritis. Southet-n Medicnl Jozïrnal70 :1 74- 1 75, 1977
Ateshian GA, Lai WM, Zhu W, Mow VC: An asymptotic solution for the contact of two biphasic cartilage layers. Journal of Biomechanics 27: 1347-1360, 1994
Ateshian GA, Wang H: A theoretical solution for the ftictionIess rolling contact of cylindrical biphasic articular cartilage iayers. Jotwnal of Bionrechanics 28: 134 1 - 1355, 1995
Athanasiou KA, Agarwal A, Dzida FJ: Comparative study of the intrinsic mechanical properties of the Iiuman acetabular and femoral head cartilage. Journal ofOrthopaedic Research 12340-349. 1994
Athanasiou KA, Agarwal A, Muffoletto A, Dzida FJ, Constantinides G,Clem M: Biomechanical properties of hip cartilage in e'cperïmentaf animal models. Ciinicai Orthopaedics and Relclted Reserch 3 1 6:254-266, 1995
Baratz ME, Fu FH, Mengato R: Meniscal tears: the effect of meniscectomy and of repair on intraarticular contact areas and stress in the human knee. A preliminary report. Arn.J.Sports.Med. 14270-275, 1986
Bergmann G, Graichen F, Rohlrnann A: In vivo messung der hüftgelenkbelastung - 1. teil: krankengymnastic. Zeitschr~lftflrOtihopadie rlnd ihre Grenzgebiere 127:672-679, 1989
Bergmann G, Graichen F, Rohlmann A: Hip joint loading during walking and running,measured in two patients, Journal of Biomechanics 26:969-990, 1993
Bergmann G, Graichen F, Rohlmann A: is staircase walking a risk for the fixation of hip implants? Jozïrnal of Biomechanics 28535-553, 1995
Bergrnann G, Graichen F, Rohlmann A, Linke H: Hip joint forces during load carrying, Clinicai Orrhopaedics und Related Research 335: t 90-301, 1997
Berpann G, Kniggendorf H, Graichen F, Rohlmann A: Influence of shoes and heel strike on the loading of hip implants. Journal of Biomechanics 28:8 17-827, 1995
S.J. Ferguson Queen's University Biomechanics of the Acetabular Labrum
[16] Biot MA: General theory of threedimensional consolidation, J-Appl.Ph_is.12: 155-164, 1941
[17] Bowen ml: Incompressible porous media models by use of the theory of mixtures- internarionaf Jozrr-nofof Engineering Science 18: 1 129-1 138, 1980
[17b] Buschmann MD, Kim YJ, Wong M, Frank E, Wunziker EB, Grodinsky AJ: Stimulation of aggrecan synthesis in cartilage expiants by cyclic loading is localized to regions of high interstitial fluid flow. Archives of Biochenrisr~and Biophysics 366: 1-7, 1999.
Byers PD, Contempomi CA, Farkas TA: A post mortem study of the hip joint including the prevalence of the features of the right side. Annals of rhe Rhezrrnntic Diseases 29: 15-3 1, 1970
Dalstra M, Huiskes R, van Erning L: Development and validation of a three-dimensional finite elernen t model of the pelvic bone, Jorrrnff[of Biomechanical Engineering 11 7272- 278, 1995
Donzelli PS, Eckstein F, Putz R, Spilker RL: Physiological joint incongruity significantly
flects the load partitioning between the solid and fluid phases of articular cartilage. - Proceeciings of the 33rd Annrral Meeting of the Orthopaedic Research Socieg, San Francisco, California, pp. 82, 1997
Dorrell JH, Catterall A: The tom acetabular labrum. Jozrrnal of Bone andJoint Stu;ary [Brj 68-B:400-403, 1986
Herberhold C, Faber S, Stammberger T, SteinIechner M, Putz R Englmeier KH, Reiser M, Eckstein F: In situ measurement of articular cartilage deformation in intact femoropatellar joints under static loading. Joirrnui ofBiorneclxznics 32: 1287-1295, 1999.
Fairban k TJ : Knee joint changes a fier meniscec tomy. Jorirnal ofBone md Joint Szrrgety [Br] 30-B:664-670, 1948
Ferguson SJ, Bryant JT, Ito K: An investigation of the function of the acetabular labrum using a poroelastic bite element model. Jorrrnal of Bone and Joint Srtrge~y[Br] 8 1- B:SUPP 1 :69, 1999
Fithian DC. Kelly MA, Mow VC: Material properties and structure-function relationships in the menisci. Clinicnl Orthopaedics and Relatei Research 33:1 9-3 1, 1990
Freeman MAR: [s collagen fatigue failure a cause of osteoarthrosis and prosthetic component miption? A hypothesis. Jorrrnal of Orîhopaeiic Research 173-8, 1999
Gardner DL: The nature and causes of osteoarthrosis. Brit. MedJ 286:4 18-424, 1983
Gibson PH, Benson MK: Congenital dislocation of the hip. Review at maturity of 147 hips treated by excision of the limbus and derotation osteotomy. Journal of Bone andJoitlt Srrrgery [Br] 64-B: 169-1 75, 1982
S.J. Fer,ouson Queen's University Biomechanics of the Acetabular Labnim
Harris WH, Bourne RB, Oh 1: Intra-articular acetabular labrum: a possible etiologïcal factor in certain cases of osteoarthritis oFthe Iiip. Jorrrnal of Bone andloint Sur-gery 61- A510-5 14, 1979
Ikeda T. Awaya G, Suzuki S, Okada Y, Tada H: Tom acetabutar labrum in young patients. Arthroscopic diaposis and treatment. Jownal of Bone and Joint Swgery [Br] 70-B:13- 16, 1988
Macirowski T, Tepic S. Mann RW: Cartilage stresses in the human hip joint. Jozrrnnl of Biomechanical Engineering 1 1 6: 1 0- 1 8, 1 994
Madreperla SA. Louwerenburg B1Mann RW, Towle CA: Induction of heat shock protein synthesis in chondrocfies at physiological temperatures. Jorrrnal of Ot-thopaedic Research 330-35, 1985
Mankin HJ: The reaction of articular cartilage to injury and osteoarthritis (tirst of hvo parts). héw England Jorrrnal of &ledicine 39 1: 1285- 1292, 1974a
Mankin HJ: The reaction of articular cartilage to injury and osteoarthritis (second of two parts). New England Jorrrnal of Medicine 29 1 :1335-1 MO, 1974b
McCarthy JC, Busconi B: The role of hip arthroscopy in the diagnosis and treatment ofhip disease. Canadian Jorrrnal of Srrrgeer y 3 8 Suppl I :S 13-1 7, 1 995
McCutchen CW: The fiictional properties of animal joints. Wear 5:1-1 7, 1962
McCutchen CW: An approximate equation for weeping lubrication, solved witli an electncal analogue. Annals ofthe Rharmatic Diseases 34:SS-90, 1975
Minerva Garcia A, Lark MW, Trippel SB, Grodzinsky AJ: Transport of tissue inhibitor of rnetalloproteinases-1 through cartilage: contribution of tluid flow and electrical migration. Jorrrnal af'Ot-rhopnedic Reseai*ch 1 6 :734-742.
Mow VC, Kuei SC, Lai WM. Armstrong CG: Biphasic creep and stress relaxation of articular cartilage in compression: theory and experiments. Journal of Biomechanical Engineering 10273-84, 1980
Mow VC, Setton LA, Guilak F. Ratcliffe A: Mechanical factors in articular cartiIage and their role in osteoarthntis, in: Osteoarthiric Disor-ders-,K Kuettner, V Go!dberg, and IL Rosemont. American Academy of Orthopaedic Surgeons, 1995
Mow VC. Soslowsky LJ: Lubricahon and wear ofjoints. In: Basic orthopaedic biornechanics, 245-292. Ed by VC Mow and WC Hayes. New York, Raven Press, 1991
Mow VC, Zhu W. Ratcliffe A: Structure and tirnction of articular cartilage and rneniscus. In: Basic Orrhopaedic Biomechank, 1, pp 143-198. Ed by VC Mow and WC Hayes. New York, Raven Press, 199 1
Poole AR: trnbalances of anabolism and catabolism of cartiIage matrix components in osteoarthritis- In: Osteoarthritic Disorders-, V Kuettner, V Goldberg, and IL Rosemont. American Academy of Orthopedic Surgeons, 1995
S.J. Ferguson Queen's University Biomechanics of the Acetabular Labnim
Prendergast PJ, van Driel WD, Kuiper JH: A cornparison of finite element codes for the solution of bip basic poroelastic pro blerns- Proceedings of the hstihftionof Mechanical Engineers. JOrrrnal of Engheering iiz Medicine [Hl 2 1 0: 13 1 - 13 6, 1996
Proctor CS, Schmidt MB, Whipple RR, Kelly MA, Mow VC: Materiai properties of the normal media1 bovine rneniscus. Journal of Orthopaedic Research 7:77 1-782, 1989
Rushfeldt PD, Mann RW, Harris WH: Impioved techniques for rneasuring in vitro the geometry and pressure distribution in the hurnan acetabulum - II. instrumented endoprosthesis measurement of articular surface pressure distribution. Jorrrnal of Biorneclianics 14:3 15-323, 198 1
Schreppers GJ, Sauren AA, Huson A: A mode1 of force transmission in the tibio-femoral contact incorporating fluid and mixtures- Proceedings of the Insrirrrtion of Mechatrical Engineers, Jorrmal of Engineering in Medicine [NJ 205:233-24 1, 199 1
Seedhom BB: Transmission of the load in the knee joint with special reference to the role of the menisci, Part 1I: expenmental results, discussion and conclusions. Engineering in Medicine 8:220-228, 1979
Seedhom BB, Takeda T, Tsubdx~M, Wright V: Mechanical factors and patellofernoral osteoarthrosis. Annals of rlze Rlzezrntaric Disemes 38:307-3 16, 1979
Shibutani N: Three-dimensional architecture of the acetabular labrurn - a scanning electron microsco pic study. Jorrrnal of the Japanese Orthopaedic Association 62:3 2 1 -3 29, 1 98 8
Simon BR: Multiphase poroelastic finite element models for soi? tissue structures. ilpplied :tferhatîics Revieiu 45: 19 1-2 18, 1992
Sol tz MA. Ateshian GA: Experimental verification and theoret ical predict ion of cartilage interstitial fluid pressurization at an impermeable contact interface in confined compression. Jorrrnai of Biomechunics 3 1 927-934, 1998a
Soltz MA, Ateshian GA: Experimental verification of cartilage fluid pressurization in confïned compression creep and stress relaxation. Proceedings of the 44th Anntral Meeting of the Ortliopaedic Resenrcll Socie@, New Orleans, Louisiana, pp. 224, 1998b
Spilker RL, Donzelli PS. Mow VC: A transversely isotropic biphasic finite element model of the meniscus. Jorrrnnl of Biomechanics 25:1037- 1045, 1992
Spilker RL, Suh JK, Mow VC: A tinite element formulation of the nonlinear biphasic model for articular cartilage and hydrated sofi tissues including straindependent permeability. In: Cornputariunal hfe~ftudsin Bioengineering., 8 1-92. Ed by RL Spilker and BR Simon, New York, American Society of Mechanical Engineers, 1988
Swann AC, Seedhom BB: The stiffhess of normal articular cartilage and the predominant acting stress levels; implications or the aetiology of osteoarthrosis- BritJRhezrm. 32: 16-25, 1993
Takechi H, Nagashima H, Ito S: intra-articular pressure of the hip joint outside and inside the limbus. Jorrrnal ofrlie Ja~aneseOrlho~aedic Association 56529-536. 1982
S.J. Ferguson Queen's University Biomechanics of the AcetabuIar Labrum
Terayama K, Takei T, Nakada K: Joint space of the hurnan knee and hip joint under a static Ioad. Engineering in Medicine 957-74. 1980
van der Voet AF, Shrive NG, Schachar NS: Numerical modelting of articu1a.r cartilage in synovial joints - poroelasticity and boundary conditions. In: Comprrrer Methods in Biomechanics and Biomedical Engineering, 200-209. Ed by J Middleton, GN Pande, and KR Williams. Swansea, IBJ Publishers, 1993
Wang LH. Soltz MA, Ateshian GA: interstitial fluid pressunzation regulates the fnctionaI response of cartilage. Proceeiiings of rhe 43rd Annz[al Meeting of the Orthopaedic Research Socief-y,San Francisco, California, pp- 83, 1997
Weber W, Weber E: Uber die mechanik der rnenschlichten gehwerkzeuge nebst der beschreibung eines versuches uber das herausfalIen des schenkelkopfes aus der pfanne im luftverdennten raum. Annalen Physik zlnd Chernie 40: 1 -1 3, 1837
Weicghtman B, Kempson GE: Load carriage. In: Adzrfr Articzrlar Carriluge, 39 1-333. Ed by MAR Freeman. London, Pitman Medicaf, 1979
Wu JZ, Herzoz W, Epstein M: Evaluation of the finite eiernent software ABAQUS for biornechanical modelling of biphasic tissues. Journal of Biomechanics 3 1: 165-169, 1998
S.J. Feraouson Queen's University Chapter Seven
The Influence of the Acetabular Labrum
on Hip Joint Cartilage Consolidation:
an Experimental Study
SJ. ~ergusonl*',J.T. ~r~ant'and K. [toi
'AO ASIF Research institute, Davos Platz, Switzerland
'~e~artrnentof Mechanical Engineering, Queen's University, Kingston, Canada Biomechanics of the Acetabular Labm
Abstract
The overall creep consolidation of six human hip joints was measured follow-inç the application of a static step load of 0.75 times bodyweight, or a dynamic sinusoidal load of 0.75 i 0.25 times bodyweight, before and after total labrum resection. The fluid pressure within the acetabular fossa - a non-contacting region of the joint - was measured. The results of the esperiment were compared to theoretical predictions of the influence of the acetabular labrum on cartilage consolidation and fluid pressurisation within the hip joint.
Fluid pressurisation kvas observed in three of the six hips. The average pressures measured were: for static loading, 54 1 * 6 1 kPa in the intact joint and 2 16 + 165 kPa following labrurn resection, for dynarnic loading, 550 * 56 kPa in the intact joint and 195
* 145 kPa following labrum resection. Up to 500 sec was required for the decay of the rneasured pressures from their peak levels.
Foliowing [abmm resection, the initiai consolidation rate was 22% greater (p =
0.02) and the final consolidation displacernent was 2 1% greater (p = 0.02). There was no si=ificant difference in the fial consolidation rate. Loading type (static vs. dynamic) had no significant effect on the measured consolidation behaviour.
The trends observed in this experiment support the predictions of the previous finite element analyses; the labrum has the potential to seal a pressunsed fluid layer within the intra-articular space of the joint and the labrum adds an extra resistance to the flow path for interstitial fluid expression from the cartilage layers, slowing cartilage consolidation.
Keywords: hip, cartilage, consolidation, pressure
S.J. Fer,ouson Qusen's University Biomechanics of the Acetabular Labrum
Introduction
It has been proposed that mechanical factors are of importance in the etiology of osteoarthrosis. Changes in the mechanical environment of the joint, or changes in the mechanical properties of the cartilage layers, can provide the necessary stimulus to initiate this degenerative procsss [23,24,27-291. In a diarthrodial joint, degeneration of the cartilage layers can be influenced by abnormalities of bone, synovial fluid, tendons, ligaments, or soft tissue accessory structures, such as menisci or Iabra [14]. In the hip joint, one such accessory structure is the acetabular labrum, a fibrocartilaginous lip attached to the perimeter of the bony acetabular socket. There is evidence that an abnormal labmm is implicated in the onset of osteoarthrosis. Clinical studies have shown that labral injuries, such as tears, can lead to changes in the joint consistent with early osteoarthrosis [ 1,10,15,I7,351.
The majority of labral tears are treated by partial or total labral resection. While the short-term results are good [13,20], the long-term effects of labrum resection are not knowm. While it is perhaps speculative to propose a direct connection between labrum pathology and the onset of osteoarthrosis, it is a hypothesis that deserves furtl~erstudy.
However, there is a scarcity of biornechanical data on the fünction of the labrum in the hip joint, with which such a mechanism could be developed and evaluated.
Several experimentai studies have demonstrated that the labmm can provide a seal against fluid tlow in and out of the intra-articular space [32,35,37]. This could improve the stability of the joint through the development of a vacuum upon joint luxation. Furthemore, this sealing hnction could enhance Lubrication rnechanisms
S.J. Fereauson Queen's University Biomechanics of the AcetabuIar Labrum
within the joint. The abiliv of the labrum to contain a pressurised layer within the joint
under load was the topic of a previous study [12] (see also Chapter Five). Using a finite-
element model of the hip joint, we predicted that the labrum could seal a fluid layer
within the joint for up to several minutes, and that this pressurised fluid would prevent
direct contact of the joint surfaces and evenly distribute the applied load across the
cartilage surfaces. However, the acetabulum presents a complex geometry and the
articulation within the joint cannot be fuIly described by such a simplified finite-element
model. A direct experimental rneasurement of fluid pressurisation within the joint would
provide valuable information to hrther explore the concept of labrum sealing.
Recently, Konrath et al [2 11 published the results of a biomechanical study of the
role of the labrum and transverse ligament in load transmission in the hip. Using pressure
sensitive films, they studied the contact pressure magnitude and distribution in the hip
before and after labrum resection. No significant differences in pressure magnitude or
distribution were observed- However, it is conceivable that the use of an intra-articular
pressure sensitive film with a thickness of approximately 250prn could alter the response
of the hip joint, when one considers the possible sealing mechanism proposed in Chapter
Five, and the relative con-mence of the joint [33]. Pressure sensitive film provides
information only about the contact pressures at the surfaces of the cartilage layers;
changes in sub-surface stress and strain levels could not be predicted fiom these
ewperiments. Furthemore, this study examined the influence of the labrum on hip joint
Ioad transmission for periods of less than one minute.
Experimental studies by Bergmann et al 13-71, using instrumented hip prostheses, have shomn that, while the maximum loads experienced by the hip joint are indeed of
S.J. Ferguson Queen's University Biomechanics of the Acetabular Labm
short duration, there is always a residual compressive force acting across the hip joint. even during sedentary periods, and that this force has, on average, a magnitude of approximately one bodyweight. In a finite-element mode1 of the hip joint, we predicted that the labrum has an influence on the consolidation of the cartilage layers of the hip
[12] (see also Chapter Six). In this analysis, the Iabm added an additional resistance to the flow of interstitial fluid expressed from the cartilage layers. Cartilage compression can occur only through the expression of interstitial fluid from the collagenous / proteogl ycan-nch solid matrix of the tissue. In the analysis, removal of the labrum increased the rate of cartilage consolidation by up to 40%, with a consequent increase in the local strain and stress ievels within the cartilage layers. While the biomechanics of the hip joint, and specifically cartilage layer consolidation and stresses, have been evaluated extensively with the use of in vitro experiments [2,8,9,19,22,30.36] and in vivo measurements [16,3 11, there have been no experimental studies of the influence of the acetabular labrum on the time-dependant consolidation of the hip joint cartilage Iayers.
Motivated by the predictions of our finite element analyses of labrurn function, the goal of this study kvas to experimentally determine the influence of the acetabular iabmrn on cartilage layer consolidation in the hip joint. Whole cadaver hip joints were loaded in a mechanical testing machine, so that the applied forces were consistent with the residual force magnitudes measured in vivo by Bergmann et al [3-71, The load was maintained for a penod of sixty minutes. The overall creep consolidation of the joint
(displacement) and the fl uid pressure within the acetabuIar fossa were measured.
S .J. Ferguson Queen's University Biomechanics of the AcetabuIar Labnim
Materials and Methods
Specimen Preparation
Four whole pelves complete with proximal femoral segments were harvested
within twenty-four hours post-mortem from fiesh human cadavera. Pelves were
obtained, and testing was conducted, with the approval of the Ethical Commission,
Department of Patliology, Cantonal Hospital of Basel, University of Basel. Radiographs
of the pelves were used to screen for evidence of advanced osteoarthrosis. The three
male and one female patients had been sixty-four, forty, forty-six and fifty years oId,
respectively, at the time of death. Pelves were harvested with al1 deep musculature,
- ligaments and joint capsules intact and then stored in airtight plastic bags and fiozen at -
20°C for tnnsportation,
Each frozen pelvis was sectioned using a butcher's bandsaw equipped with a
specially constructed alignment fiame (Figure 7.1). The first cutting plane was a
transverse plane through the ilium, parallel to a line joining the iliac crests and
perpendicular to the coronal plane, which \vas defined by the anterior superior iliac
spines and pubic symphysis. The second cutting plane was through the midline of the
sacrum, perpendicular to both the coronal pIane and the previous transverse cutting plane.
Two herni-pelves were selected for a separate anatomical study. Hemi-pelves were
stored at -20°C until the time of testing.
S.J. Ferguson Queen's University Biomechanics of the Acetabular Labrum
Figure 7.1 : Whole pelves were sectioned while fiozen usinp an orthogonal cutting guide on a bandsaw. The first cutting plane was a transverse plane through the ilium, parallel to a line joining the iliac crests and perpendicular to the coronal plane (left). The second cutting pIane was through the midline of the sacrum, perpendicular to both the coronal plane and the previous transverse cutting plane (right). Plastic bones used for illustrative purposes.
Figure 7.2: Using the three orthogonal cutting pIanes for reference, the hemipehis was rnounted in a plexiglass hme(left). Two cortical bone screws and one Steinmann pin used for fixation. With the pelvis finnly held, the sacrum couId be cut away. The fiame was then mounted to a vertical slide for PMMA potting of the acetabulum (right). The rotation of the hune produced the desired 16" medial ali~gunentof the joint reaction force during testing.
S.J. Ferguson Queen's University Biomechanics of the AcetabuIar Labrum
Pnor to testing, the hemi-pelvis was thawed and cleaned of al1 soff tissue, escept
for the joint capsule and labrum. Using the three orthogonal cutting planes for retèrence,
the pelvis was mounted in an alignment fiame with one rotational degree of freedom in
the coronal plane. The sacrum was cut away and the hemi-pelvis was rotated and potted
in polymethyl-methacrylate (PMMA) cernent, such that the resultant joint reaction force was aligned 16" medial of vertical, relative to the acetabulum (Figure 7-21, based on the
Ioading studies of Bergmann er al [3-71.
A miniature pressure transducer (Mode1 060S, Precision Measurement Co., Ann
Arbor, Michigan, USA), with dimensions of 1.00mm diarneter x 0.30mm diameter thickness was inserted into the joint (Figure 7.3).
Figure 7.3: Placement of the pressure transducer. The pressure transducer was placed within the fat pad of the acetabular fossa The fat pad itself is very sol? and unable to develop significant levels of solid stress within the tissue. Therefore, the pressure measurements obtained in the fossa shouid reflect the pressurisation of the fluid within the intra-articular joint space.
S .J. Ferguson Queen's University Biomechanics of the Acetabular Labnim
A 2.Ornrn diameter trochar was inserted through the soft tissue of the acetabular notch,
under the transverse acetabular ligament, into the fat pad of the fossa. The transducer
was ptaced within the fat pad of the fossa to provide a measurement of fluid
pressurisation (if any) within the intra-articular space of the joint, rather than the direct
contact pressure between the articular cartilage surfaces. The pressure transducer was
inserted through the trochar into the fossa and held in place as the trochar was withdrawn.
Soft tissue sealing around the transducer lead wires \vas supplemented with a "purse sûing" suture and cyanoacrylate glue.
Testing Procedure
The potted hemi-pelvis was positioned in an environmental chamber mounted on a servo hydraulic testing frame (MTS Bionix, MTS Systems Corporation, Eden Prairie,
MN ,USA). The acetabulum was visuaIly aligned over the centreline of the testing fiarne and then bolted in place. The proximal portion of the femur was aligned vertically in the sasgital plane, with 9O of adduction relative to the acetabulum and with no interna1 or external rotation. The proximal femoral shaft was potted in sitzr with PMMA in an adjustable holding fixture attached to the moving hydraulic actuator of the testing frarne.
The environmental chamber itself was mounted on a baI1-bearing table with two linear and one rotational degrees of fkeedorn to allow free centring of the joint and to eliminate any lateral forces (Figure 7.4).
S JIFeraouson Queen's University Biomechanics of the Acetabular Labmrn
Figure 7.3: Testing apparatus mounted in the MTS Bionix servo-hydraulic test frame. The acetabuhm is fised in Pm,with bone screws for additional stability. The kmur is fixed in PMMA with O" of internailextemal rotation, Oa tlexiodextention and 9" of adduction relative to the pelvis. The environmental chamber is filled with 0.9% PBS solution, heated to 37°C by an immersion heater. The resuitant joint reaction force is angled 16" medially From vertical, relative to the acetabulurn. The entire testing apparatus rests on a ball-bearing table on top of the MTS load cell, allowing the joint to self-centre during loading.
Reflective markers for use with a video-based motion analysis system
(MacReflex, Qualisys AB,Goteborg, Sweden) were attached to the joint. Two rnarker chsters (each w-ith five retlective spherical markers, 6mm in diameter, arranged in a 25 x
25 mm cross pattern) were mounted on 3Smm diameter steel Kirschner wires attached to the two joint components. One Kirschner uire was driven into the femoral head through a cIearance hole fiom the lateral aspect of the greater trochanter through the femoral neck and the second Kirschner wire \vas driven into the ossis pubis beside the acetabular rim.
The Kirschner wires were positioned such that the marker clusters were clearly visible in
S.J. Fer,ouson Queen's University Biornechanics of the Acetabular Labnun
the field of view of the wovideo cameras. which were located 2rn away from and lm above the specimen, with a camera separation of 'm.
Following specimen alignment, the joint capsule \vas excised. The environmental chamber was filled with 0.9% phosphate buffered saline solution (pH 7.4) and the temperature of the bath was maintained at 37°C (* 2°C) with an immersion heater. The joint was allowed to equilibrate to the bath temperature. The position of the hydraulic actuator was adjusted until there was no net tensile or compressive force applied to the joint, and this zero position was recorded.
A total of five creep-consoIidation/recoverycycles were performed on each joint: one preconditioning cycle with a static load, one cycle with a static load and one cycle with a dynamic load applied to the intact joint, and one cycle with a static load and one cycle with a dynamic load applied following labrum excision. Here, static loading retèrs to a constant load of 0.75 x bodyweight and dynarnic loading refers to the superposition of a sinusoidally varying Ioad onto a constant load (0.75 * 0.25 x bodyweight, IHz frequency). Each cycle was performed in the same manner. Prior to each loading cycle, a preload of SON was applied and held for one minute to allow the joint to seat itself.
Subsequentty, the chosen load was applied by a ramp over ten seconds and maintained for one hour. The tirne constant for fluid expression fkom the cartilage layers is suffrciently long that the ten second ramp loading was, in effect, a step load. Following the consolidation period, the joint was unloaded and a small tensile Ioad (25N) was maintained across the joint to allow return of fluid into the cartilage layers during a one- hour recovery period. Following the recovery period, the actuator was retumed to its zero-load position and the next consolidation/recoverycycle was performed.
S.J. Ferguson Queen's University Biomechanics of the Acetabular Labrum
------
During the creep-consolidation penod of the testing, the displacement of the hydraulic actuator, and thus the overall consolidation of the joint, was measured by the linear variabIe differential transformer (LVDT) of the testing frame and recorded at
10Hz. Pressure data were collected on a personal computer at a sampling rate of IOOHz, and filtered during collection using a LOHz low-pass filter to coincide with the raw displacement data. Due to the length of the testing, video data from the motion analysis system could be recorded at a maximum of 2Hz to the video processors, then was subsequently downloaded and analysed on a persona1 computer for the reconstruction of three-dimensional spatial coordinates from the stereo video images.
Data Analysis
Al1 data collected during testing were subsequently tlltered to 0.2Hz for consistency before analysis. The theconstant for the cartilage Iayers of the joint is sufFrcientIy iong that the displacement and pressure data values meûsured dunng the consolidation tests were not influenced by tiltering. Data Eom the ramp loading period and subsequent ten seconds of consolidation was also examined before fiItering for determination of peak pressure and initial displacement after loading without the possible errors in the initial transient response introduced by filtering.
The overal1 consolidation of the joint was measured by the displacement of the hydraulic actuator of the testing fiame. This provided a measure of the overall displacement of the joint components, including the bone and any flexibility of the testing hmeand fixtures. The three-dimensional video motion data from the two joint components was used to calculate the relative angular orientation (cardan angles) of the
S.J. Ferguson Queen's University Biomechanics of the Acetabular Labmm
femur and the acetabulum in the global CO-ordinatesystem, defined within the testing fiame. The resolution of the motion analysis data (approximately SOpm for each marker, * 100pfor relative motion of nvo marker clusters) was not sufficient to directly rneasure the relative consolidation of the two subchondral bone surfaces.
The final consolidation displacement of the joint was calculated as the average of the last 100 seconds of displacement data for the test. The initial consolidation rate of the joint was calculated by a linear regression of the first LOO seconds of consolidation dispIacement data after the end of the ramp loading period. The long-term consolidation rate was calculated by a liner regression of the last 1000 seconds of the consolidation displacement data. To eliminate the influence of the load magnitude on different sized joints, consolidation displacements and rates were normalised by dividing by the applied load. Measured and calculated consolidation displacement parameters were compared for di fferences between loading regi mes (static and dynamic) and for di fferences be fore and after labrum resection usin,o an analysis of variance - general linear model, as implemented in SPSS 9.0 (SPSS hc., Chicago, USA). The SPSS general linear model is a logistic regession model based on a block design. which tests for differences within and between blocks (hips). Differences in outcome variables were tested for significance against the individual independent variables, as well as for any interaction behveen independent variab les, fol lowing graphical tests for nomality and homogeneity of the distribution of the sample populations. A significance level of p < 0.05 was selected.
S.J. Ferguson Queen's University Biomechanics of the Acetabular Labrum
-
For the direct cornparison of temporal data tiom the pressure measurements and
consolidation displacement measurements, the measured data \vas fit to single- and
double-exponential regressions curves, using a Ieast-square, non-linear regession curve-
fitting procedure with the solver feature of Excel (Excel97, Microsof? Corp,, Seattle,
USA).
Skhips were tested f?om four donors. Tlxee donors were male (four hips) and
one donor was female (two hips). Post-testing visual examination confirmed the pre-
testing radiographie evaluation; al1 joints were in good condition, with hl1 coverage of
both articular surfaces? normal cartilage appearance, no osteophyte development, and a normal, intact labrum.
Pressure Measurements
Pressure readings were taken during manual positioning of one of these joints and demonstrated that a negative pressure of up to -100 kPa was developed during manual joint subluxation. During the consolidation expenments, pressure measurements were recorded in five of the six joint tested. The pressure transducer was damaged following insertion in the final test and no pressure data was obtained. Three of the five joints (the gith hip tiom specimen #O58 and the left and right hips From specimen #170) exhibited a measurable non-zero pressure within the fossa following loading (Figures 7.5 - 7.10).
S.J. Feraouson Queen's University Biomechanics of the AcetabuIar Labrum
A98.058 - Static Load
. - -labrum -no labrum
time (sec)
Figure 7.5: Fluid pressure recorded in the acetabular fossa (78kg male. right hip) following application of 57ON step load. The esperiment was repeated folloviing total labrum excision. Pressure magnitude and duration were substantially lower-
A98.058 - Dynamic Load
time (sec)
Figure 7.6: Fluid pressure recorded in the acetabular fossa (78kg male. right hip) following application of 57ON* 190N sinusoidal dynamic load. The experiment was repeated following total labrum excision.
S-J. Ferguson Queen's University Biomechanics of the Acetabutar Labm
A98.170 - Static Load
tirne (sec)
Figre 7.7: Fluid pressure recorded in the acetabular fossa (50kg fernale, right hip) following application of 3ïON step load. The experiment was repeated following total labnim excision.
A98.170 - Dynamic Load
------labrum -no labnirn
time (sec)
Figure 7.8: Fluid pressure recorded in the acetabular fossa (50kg female, right hip) following application of 3 7ONk 120N sinus0 idal dparnic load. The experirnent was repeated following totaI labrum excision.
S.J. Feraouson Queen's University Biomechanics of the Acetabular Labrum
A98.170L - Static Load
tirne (sec)
Figure 7.9: Fluid pressure recorded in the acetabular fossa (50kg female, lefi hip) follow-ing application of370N step load. The experiment was repeated foIlowing total labrum excision. The pressure recordings decayed fiorn their peak to a minimum after approximatdy 400-500s, but then increased-
A98.170L - Dynamic Load
tirne (sec)
Figure 7.10: Fluid pressure recorded in the acetabular fossa (50kg female, left hip) following application of 37ON+ I20N sinusoidd dynarnic load. The experiment was repeated following total labrum excision.
S .J . Ferguson Queen's University Biomechanics of the Acetabular Labm
In these three joints, peak pressure readings were obtained immediately following loading, afier which the pressure readings decreased exponentially with increasing time.
Over 500 s was required for the decay of the measured pressure tiom its peak reading. in general, there was no difference in the pressure readings for static versus dynamic loading, but there was a large difference bebveen the pressure readings before and after labrum excision. The maximum pressures recorded in these three hips were, on average:
541 * 6 1 kPa during static loading with an intact labmm, 2 16 i 165 kPa during static loading following labnim removal, 550 * 56 kPa during dynamic loading with an intact labrum and 195 * 145 kPa during dynamic loading following labrum removal. Analysis of the near exponential decay of each pressure reading fiom the peak reading revealed that the time constant of this decay was up to 94% greater in the intact joint. In one specimen, the measured pressure decreased to almost zero, in the second, the pressure reading stabilised at a non-zero equilibrium value and in the third, the pressure gradually increased again afier reaching a minimum at approximately 400 S.
Displacement Measurements
Ali joints exhibited an initial linear displacement of the femur into the acetabulum upon load application as the joint seated itself, followed by a non-linear consolidation of the cartilage Iayers, with the femur gradually settling deeper into the acetabulum with increasing time. After 3 600s of loading, the joint had still not reached a steady-state displacement. For analysis, the initiai linear portion ofthe displacement curve was excIuded, as there was a degree of uncertainty in the initial position of the joint foIIowing
S-J, Ferguson Queen's University Biomechanics of the Acetabular Labrum
the recovery period. Hence, the overall displacement of the joint components at the end of the ten second loading ramp was taken as the initial displacement for subsequent calculations of the non-linear creep consolidation displacement.
Consolidation displacement measurements calculated from the spatial coordinates of the video markers were not suficiently accurate with respect to the actual magnitude of the joint displacements. Therefore, displacement measurements €rom the MTS actuator LVDT were used to characterise the overall joint consolidation. The relative change in angular orientation of the two joint components (acetabulum and Femur) during testing was not more than Io about the global X, Y or Z axes, as calculated from the video motion data. The average change in angular position of the joint components was
OS0 over the duration of the experiments.
Representative displacement data rneasured by the LVDT of the hydraulic actuator, and hence the overall consolidation of the joint, are presented in Figures 7.1 1 -
7-14, and summarised in Table 7.1. The initial consolidation rate following labrum resection ( 1-771 * 0.145 pm/s-N) was significantly higher (p = 0.02) than with the intact labmm ( 1.454 i 0.095 pm/s-N). There was no significant difference (p = 0.27) in initial consolidation rate with respect to loading type. The final consolidation rate atter 3600 seconds was approxirnately 25 times slower than the initial consolidation rate. There was no sipificant difference in final consolidation rate following labrum resection (p = 0.57) or with respect to loading type (p = 0.76). The final normalised consolidation displacement following labrurn resection (0.748 * 0.050 prn/N) was significantly greater
(p = 0.02) than with an intact labrum (0.6 18 0.027 pm/N). There was no significant change in the final consolidation displacement (p = 0.38) with respect to loading type.
S.J. Ferguson Queen's University Biomechanics of the Acetabular Labm
A98.058 - S tatic Load
time (sec)
Figure 7.1 1 : Displacernent of the hydraulic actuator following application of a step Ioad (3/4 bodytveight - 78kg male, right hip). Downward dispIacernent of the actuator corresponds to motion of the fernur into the acetabulum and cornpaction of the cartilage layers, Measurements were repeated following labmm excision.
A98.058 - Dynarnic Load
time (sec)
Figure 7.12: Downward displacernent of the hydraulic actuator following application of a dynamic load (34bodytveight k 1/4 bod,vweight sinusoïdal component - 78kg male, right hip).
S.J. Ferguson Queen's University Biomechanics of the AcetabuIar Labrum
A98.227L - Static Load
-no labnim
time (sec)
Figure 7.13: Displacement of the hydraulic actuator following application of a step load (34 bodyweight - 68kg male. lefi hip). Downward displacement of the actuator corresponds to motion of the îèmur into the acetabulum and compaction of the cartilage layers, Measurements were repeated fol lowing labrum excision.
A98.227L - Dynamic Load
labnim 1 -na labrurn
time (sec)
Figure 7.14: Downward displacement of the hydraulic actuator following the application of a dynamic load (Y4 bodyweight * 1/4 bodyweight sinusoidal component - 68 kg male, lefi hip).
S.J. Ferguson Queen's University Biomechanics of the Acetabular Labm
Table 7-1 : Summary of consolidation displacernent measurements. Data is nomalised to the applied load,
Normal ised Initial Norrnalised Final Norrnalised Final Group Consolidation Consolidation Displacement Rate(pm/s-N)x10" Rate(pm/s-Mx10" (P~W
intact Labrum 1.454 i0.095* 6.477 * 0.41 O 0.6 18 = O-037*
Resected Labrum 1.77 1 k O. 135* 6.669 * 0.566 0.748 k 0.050" Static Loading 1.685 = 0.134 6.52 1 * 0.479 0.706 * 0.037 Dynamic Loading 1.539 = O, 126 6.625 * 0.508 0.660 * 0.05 I
Ieast squares mean = standard error * p = 0.03
Cornparison of the pressure recordings, Born the three hips which produced non- zero measurements. with the displacement measurements from the same hips showed a correlation benveen the changes in the rate of pressure decay and the initial consolidation displacement rate. This cornparison is presented in Table 7.2. A cornparison of the pressure decay time constants with the calculated time constants of the overall joint consolidation for one specimen is presented in Table 7.3.
S.J- Fer-won Queen's University Biomechanics of the Acetabular Labrum
Table 7.2 :Comparison of pressure decay time constant and initial consolidation rate. Results are presented as the ratio of values from tests before and afler Iabrum resection (Iabrum :no Iabmm)
Test Nomalised initial Pressure Decay joinH (side) load Consolidation Rate Ebtio Time Constant Ratio (prn/s-N) n 1O-' r (SI*
058 (nght) static 1-46 :2.5 1 1 : 1-72 169 : 103 1.64 : 1
058 (right) dynarnic 1.20 : 1-94 1 : 1.62 183 : 116 1.57 : 1
170 (right) static 1.14 : 3-35 1 : 1.97 189 :97 1.94 : 1
170 (right) dynarnic 2.00 :223 1 : 1-12 1 14 : 95 1.19 : 1
170L (Iefl) static 1.83 :3.33 1 : 1.28 91 :77 1.18 : 1
170L (lefi) dynamic 1.83 :3-05 1 : 1-13 59 : 57 1 .O6 : 1
Table 7.3 : Comparison of pressure decay tirne constant and consolidation disptacement time constants for one specimen (A98.058 - right Iiip)
Pressure Decay Consolidation Consolidation Consolidation Test Time Constant Time Constant Time Constant Time Constant r (SI* ri (SI** -c2 (s)*** 5; (SI*** static - labmm 169 static - no labrum 103 dynamic - labrum 182 dynamic - no labrum 116
Discussion
S.J. Fer,pson Queen's University A variety of methods have been used in previous experimental studies to determine the changes in cartilage layer thickness in loaded joints: magnetic resonance
imaging (MRI) [ 1 1,181, instrurnented endoprostheses [22,30,34], roentgenographic techniques [2], and direct instrumentation within the joint [9]. Each method had its
limitations: poor displacement resolution, low sampling frequency, or disruption of the natural joint. Our intention had been to use a three-dimensional video motion analysis systern to provide a non-invasive measurement of the relative displacement of the two joint components during testing. However, while the resolution of such systems is theoretically adequate for such measurements, our experîence was that errors introduced during the measurement and analysis of the consolidation of actual joint specimens were too large to consider this technique. A discussion of this technique and its limitations is presented in Appendix E.
Because of the lack of a satisfactory method for directly measuring cartilage layer consolidation in the intact joint, we chose to use the overall displacement ofthe joint within the testing machine to detennine the influence of the labrum on cartilage consolidation. implicit in this choice of displacement measurement are assumptions about the relative response of the bone and soft tissues of the joint to loading, and the overall stiffness of the testing apparatus. In order to eliminate some of these unknown effects fiom the analysis of the joint consolidation, only the non-Iinear time-dependent displacement afier load application has been considered. This elirninated any motion of the testing apparatus upon loading, and the initial elastic deformation of the joint components in response to the applied load. Based on published values of bone stiffhess and permeability, the creep time constant of bone is several orders of magnitude less than
S.J. Ferguson Queen's University Biomechanics of the Acetabular Labrurn
that of the cartilage layers and the labrum. Therefore. it was assumed that the contribution of bone creep to the overall displacement of the joint following loading is negligible. Based on the initial consolidation rate measured during testing and the sampling fiequency of the readings fkom the LVDT on the MTS testing frame, uncertainty in the displacement measurement chosen as the zero point follotving load application would not exceed 5pm.
Pressures recorded during these experiments were not entirely consistent.
Pressure data was not recorded in one of the six hips due to a technical failure of the instrumentation after implantation. Non-zero pressures were recorded in three of the five remaining hips during testing, For these three hips, the pressures recorded during static or dynamic loading in the intact hip joints were characterised by a high peak pressure
(400 - 600 kPa) which decayed slowly over the following 400 - 1O00 seconds, It is interesting to note that the peak pressure measured in the fossa is much higher than the peak systolic pressure of the circulatory system (approximately 16 kPa), In one hip, the pressure decayed to almost zero. In a second hip, the pressure decayed to an equilibrium value of over 200 kPa. in the third hip ( 170L), the pressure began to rise steadily after reaching a minimum value. After resection of the Iabrum, the peak pressure reached was much lower, except for hip 170L, Care was taken to position the pressure transducer in the middle of the fat pad in the acetabular fossa- However, the pressure transducer in hip l7OL was pushed too far during insertion and upon joint disarticulation following testing, we found that the transducer was placed at the edge of the acetabular cartilage surface
(Figure 7.15). It is likely that the transducer was pressed between the opposing cartilage surfaces at some point during testing as the joint consolidated and fluid was expressed
S.J. Ferguson Queen's University Biomechanics of the Acetabular Labrum
-- - from the intra-articular cartilage space. As this joint consolidated, the increasing size of the cartilage contact area within the joint, and resulting increase in contact pressure with time at the edge of the cartilage surfaces, could explain the gradua1 increase in the pressure measurements in this joint. In the other bvo joints, the pressure transducer was found to be deep inside the acetabular fossa when the joints were opened and exarnined afier testing. It is unlikely that the measurements obtained from these joints reflect the direct contact of cartilage surfaces within the joint. The fat pad of the fossa itself is very sofi and unabIe to develop significant levels of solid stress within the tissue. We assumed, therefore, that the pressure measurements obtained in these two joints, and in the early period of loading in hip 170L, reflect the pressurisation of the fluid within the fossa.
Figure 7.15: Pressure transducer placement, Normal placement of the transducer was within the fat pad of the fossa (lefi - arrow), In one hip (170L)the transducer was inserted beyond the deepest point of the fossa. just to the edge of the articular cartilage of the facies Ztrmfu (right)
Decreases in the peak pressures measured following labnim resection indicate that the labrum has an influence on this pressurisation, and that therefore the ffuid within the
S.J . Ferguson Queen's University Biomechanics of the Acetabular Labm
fossa must communicate with the labrum, implying the presence of a pressurised fluid layer between the two cartilaze surfaces. Subjective evaluation of the "tightness" of the labrurn seaI around the femoral head supports this interpretation of the results. The three joints which produced a measurable non-zero pressure under load had tightly-fitting labra which closely foilowed the cartilage sufiace of the Femoral head. In the hips which did not produce pressure rneasurements, the labra were loose fitting, with a visible gap in places between the femoral head and the labrum. During joint positionhg and loading, fluid could be observed flowing past the labrurn in these hips. The sealing function predicted in Chapter Five relied on both an advantageous geometry of the joint and suEcient mechanical stiffness of the labrum tissue itself. A siightly incongruent joint, with initial contact between the fernoml head and the acetabular rim, would enhance seaiing, whereas the opposite case, a slightly undersized femoral head within the acetabulum, would reduce the eftèctiveness of this seaI, Age-related changes to the labrum [32b] may also be responsible for the relatively leaky seal, or complete lack of sealing, in some joints.
The time response of cartilage consolidation was highly non-linear. The initial consolidation rate was approximateIy 25 times greater than the consolidation rate after
3600 seconds of loading. Kowever, even after 3600 seconds the joint had not reached an equilibrium state. The greatest change in cartilage consolidation rate occurred within the first 500 seconds. Macirowski et al [22] observed an abrupt change in the consolidation rate of the cartilage layers of the acetabulum after approximately 120 seconds. or 60pm of displacement, when loaded by a spherical meral endo-prosthesis. They concluded that the development of an intra-articular "seal" of the gap between the two surfaces was
S.J. Ferguson Queen's University Biornechanics of the Acetabufar Labrurn
responsible for this non-linear behaviour, and that this seal resulted fiom the contact and
flattening of the irregularities of the cartilage layers. Furthermore, they proposed that
this sealing effect occurred over the entire surface of the joint and was responsible for the
considerably decreased consolidation rate. However, RushfeIdt et ai [30] demonstrated
that the consolidation behaviour of the acetabular cartilage layer, using the same endo-
prosthesis as Macirowski et ai, was greatly influenced by the fit of the spherical
prosthesis bal1 in the acetabular socket. While not specifically stated, it would appear,
from the illustrations which accompany both these studies, that the labrum \vas resected
prior to testing. It is possible that the human hip, together with the acetabutar labrurn,
initially possesses the characteristics of an elastohydrodynarnic bearing, with Ioad
supported by a fluid film of uniform pressure and thickness- surrounded at its perimeter
by the labrum, which provides a seal against fluid expression from this film. The finite
element analysis described in Chapter Five provided sorne evidence that the labnim could
function as such a seal, for a period of up to several minutes- Incon_auity of the hip joint,
with initial contact between the femoral head and the acetabular perimeter, enhances this
mechanism. The expenmentai studies of Takechi et al [32] and Terayarna et al [35] also
Iend support to the hypothesis that, at Ieast initially, the labrum can maintain a
pressurised fluid layer within the joint. in the current study, the initial consolidation rate
\vas 22% faster following labrum resection than with the intact labrum. The greatest differences in consolidation behaviour before and afier labrum resection were observed in the hips which produced pressure readings during testing, and which subjectively had a better fit between the labrum and fernoral head, Furthermore, it was interesting to note
SJ, Ferguson Queen's University Biornechanics of the Acetabular Labrum
that the changes observed in the initial consolidation rate of the joints following Iabrum excision corresponded to changes in the time constant of the pressure decay.
Previously, finite element models have been used to investigate the influence of the labnxm on long-tenn cartilage layer consolidation in the hip joint (see Chapter Six).
These models predicted that the cartilage layer consolidation rate would be up to 40% faster following resection of the labrum. Furthermore, the analyses predicted that the
Merconsolidation rate would be accompanied by changes in the contact pressure magnitude and distribution within the joint, and higher levels of subsurface stress in the cartilage layers. These models examined the response of the joint to a vertical load.
Subsequently, these models have been re-analysed, changing the direction of the joint reaction force to 16" medial OF vertical in the acetabulum, to simulate the conditions of the experiment. The overall consolidation rate in this case was up to 13% faster following labrum resection. The magnitude of cartilage layer consolidation predicted by the mode1 (250 - 350 pm) was consistent with the overail joint consolidation measured in the experiments. The experimental measurements showed that the initial consolidation rate was, on average, 22% higher foliowing labt-um resection, that the total consolidation displacement was 2 1% higher, but that the final consolidation rate was not signitïcantly different. Therefore, the hypothesis of faster joint consolidation following labnim resection, as predicted by the finite element models, has been supported by the results of these experiments. The increase in joint consolidation rate measured in the experiments
(22%) was consistent with that predicted by the rnodels (13 - 40%). It would appear that the influence of the labrum on cartilage layer consolidation over long periods of time depends on the orientation of the applied load, due to the position of the labrurn in
S.J . Ferguson Queen's University Biomechanics of the Acetabular Labrum
relationship to the loading direction. For more vertically oriented loading, the affect of
the increased resistance provided by the labrurn to fluid expression fiom the cartilage
layers is more pronounced.
The results of these hip joint consolidation experiments have indicated that the
labrurn has its greatest influence on the response of the hip joint to load in the first 500 -
1000 seconds aker loading. Pressure measurernents taken during the consolidation
experiments imply that these differences rnay be partly due to the ability of the labmm to
seal a pressurised layer of fluid within the joint. Previous cornputer analyses have
predicted that such a fluid layer would be depleted afier approximately 200 seconds. The
differences observed in these experiments rnay also be due to the increased resistance
presented by the labrurn to fluid expression from the cartilage layers, as has also been
predicted by compter models. The consolidation displacement data were best described
by a mathematical expression which included twfo exponential components, one with a
short tirne constant and one with a long time constant, plus a linear term. The initial
rapid consolidation rnay be reIated to the rate of fluid expression from between the hvo cartilage surtàces of the joint, while the long term consolidation, governed by a slower exponential decay plus a linear term, rnay be descriptive ofthe inherent consolidation behaviour of the cartilage Layers themselves. Decreases in the fast time constant rnay reflect the Iack of a labrurn seal against fluid layer depletion, while changes in the slow time constant rnay reflect the decreased resistance to fluid flow following labtum resection. The longer time constants associated with dynamic loading rnay indicate that these sealing mechanisms are more efficient under cyclic loading. Most likely it is a combination of the two rnechanisms, but the correlation between the increased rate of
S.J, Ferguson Queen's University Biomechanics of the AcetabuIar Labrum
pressure decay following labrum resection with the increased rate ofjoint consolidation
provides a convincing argument for the dominant role of the initial sealing of the joint
space by the labrum in the overall consolidation behaviour ofthe joint.
Fiwre 7.16: A deep fissure was noted in one hip (227R) at the junction of the labrum and ? cartilage (arrow - labrum resected). There was some erosion of the cartilage. Othenvise, the cartiiage surfaces of the acetabulum and femur were normal (smooth,glossy, white)
Al1 hips were examined following testing for signs of cartilage or labrum degeneration. Overall, the condition of the cartilage surfaces in al1 hip joints kvas very good, with thick, uniform cartilage layers and no signs of fibrillation. In one hip, there L. was a fissure observed at the interface between the labrum and the acetabular cartilage, in the supetior region of the acetabulum. This fissure was approximately 2Omm long and the edges of the cartilage surface were eroded in this area (Fibure 7.16). The labrum was
S.J. Ferguson Queen's University Biomechanics of the AcetabuIar Labmrn
not tom and appeared to be normal; however, in this hip, the fit between the labrum and the femoral head \vas Iooser than in the other hips. Loss of mechanical support from the
labrum in this highly loaded region of the acetabulum rnay have resulted in higher stresses in the adjoining cartilage and consequently, darnage to the tissue.
The previous computer models have shown that, for changes in overall joint consolidation consistent with the experimental results, local stress and strain levels within the cartilase layers of the hip joint may be higher. These higher stress and strain levels within the joint may lead to frictional wear of the cartilage surface, or fatigue failure of the collagenous matrix of the cartilage [26], resulting in the fibrillation and delamination that has been observed in hip joints with labrum damage.
S J-Ferguson Queen's University Biomechanics of the Acetabular Labrum
References
Altenberg AR: AcetabuIar labrum tears: a cause of hip pain and degenerative arthritis. Sorrthem :t/edical Jorirnal70 :1 74- 1 75, t 977
Armstrong CG, Bahrani AS, Gardner MA: In-Vitro measurement of articular cartilage deformations in the intact human hip joint under load. Jorrt-nal of Bone andJoint Sru-gery 6 1-A:744-755, 1979
Ber-mam G, Graichen F, Rohtmann A: In vivo messung der hüftgelenkbelastung - 1. teil: krankengymnastic. Zeirschrfrfiir Orrhopndie und ihre Greul-gebiete 127:672-679, 1989
Bergmann G, Graichen F, Rohimann A: Hip joint Loading dtiring walking and running,measured in two patients. Jorrt-nal of Biornechanics 26:969-990, 1993
Bergrmm G, Graichen F, Rohlmann A: Is staircase walking a risk For the fixation OFhip implants? Jorn-na1 of Biornechanics 28535-553, 1995
Bergmann G, Graiclien F, Rohlmann A, Linke H: Hip joint forces during load carrying- Clinicai Orthopaedics and Relared Research 335: 190-20 I , 1997
Bergmann G, Kniggendorf H, Graichen F, RohImann A: Influence of shoes and heel strïke on the Ioading of hip implants. Jorrrnal of Biomechanics 28:8 17-837, 1995
Brown TD, Shaw DT: in vitro contact stress distributions in the natural hurnan hip. Jorwna/ of Biomeclrania- 1 633 73-384, 1983
Day WH, Swanson SA. Freeman MA: Contact pressures in the loaded human cadaver hip. Jorrrnal of Bone andJoinr Szrrgety 57-A:302-3 13, 1975
Dorrell JH, Catterall A: The tom acetabuIar labrum. Jorrrnul of Bone anJJoinf Srrrgety [Br] 68-B:400-403, 1986
Eckstein F. Tieschky M. Faber SC, Haubner M, Kolem H, Englmeier K-H, Reiser M: E ffect of physical exercise on cartilage volume and thickness in vivo: MR imaging study. Rndiologv 207243-248, 1998
Ferguson SJ, Bryant JT, tto K: An investigation of the function of the acetabular labrum using a poroelastic finite element model, Jownal of Bone and Joint Srtrgery [Br] 8 1 - B:SUPP 1:69, 1999
Fitzgerald RH: AcetabuIar Iabrum tears. Cfinical Orrlzopaedics and Relared Research 3 1 1:60-68, 1995
Gardner DL: The nature and causes of osteoarthrosis. Brir. Med-J. 286:418-424, 1983
Gibson PH, Benson MK: Congenital dislocation of the hip. Review at maturity of 147 hips treated by excision of the Sibus and derotation osteotomy. Jozrt-nul of Bone andJoi~ Srrrgery [Br] 64-B: 169- 1 75, 1 982
S.J. Fermauson Queen's University Biomechanics of the Acetabular Labm
Givens-Heiss DL, Krebs DE, Riley PO, Strïckiand EM, Fares M, Hodge WA, Mann RW: In vivo acetabular contact pressures during rehabilitation, part II: postacute phase. PhysicnZ Therapv 72700-705, 1992
Harris WH, Bourne RB, Oh 1: Intra-articular acetabular labrurn: a possible etiological factor in certain cases of osteoarthritis of the hip. Joirrnal of Bone andhint Srrt-gery 61- A:5 10-5 14, 1979
Herberhold C, Stammberger T. Faber S, Putz R, Englrneier K-H, Reiser M, Eckstein F: An Mr-based technique for quantifying the deformation of articular cartilage during mechanical loading in the intact cadaver joint- hfagnetic Resonance in Medicine 399343- 850, 1998
Hodge WA, Carlson KL, Fijan RS, Mann RW: Contact pressures tiom an instrumented hip prosthesis- Joru-nul of Bone nndJoint Sur-gery 7 1-A: 1378-1386, 1989
ikeda T, Awaya G, Suzuki S. Okada Y, Tada H: Tom acetabular labrum in young patients. Arthroscopie diagnosis and treatment. Jozrrnal of Bone andJoinr Sttrgery [Bd 70-B: 13- 16, 1988
Konrath GA, Harnel AJ, Olson SA, Bay B, Sharkey NA: The role of the acetabuIar labrum and the transverse acetabular ligament in load transmission in the hip. Joicrnal ofBone und Joint Sirrgery 80-A: 178 1 - 1787, 1998
Macirowski T. Tepic S. Mann RW: Cartilage stresses in the human hip joint. Jorrri7al of Biom echaizical Engineering 11 6:1 0- 18, 1994
Mankin HJ: The reaction ofarticdar cartilage to injury and osteoarthritis (first of hvo parts). New EnglrrndJorrrnal of Medicine 29 1:1285-1292, 1974a
Mankin W: The reaction of articuIar cartilage to injury and osteoarthritis (second of two parts). ~VewEnghndJorrrnal of Medicine 29 1: 1335-1340, 1974b
McCarthy JC. Busconi B: The role of hip arthroscopy in the diagnosis and treatment of hip disease. Canadian Jorrrnal of Strrgery 38 SuppI 1 :S 13-1 7, 1995
Mow VC, Soslowslq LJ: Lubtication and wear of joints- In: Basic orthopaedic bionrechanics, 345-293. Ed by VC Mow and WC Hayes, New York, Raven Press, 1991
Mow VC, Zhu W, RatcIiffe A: Structure and function of articular cartilage and rneniscus. In: Basic Orthopaedic Bionzeclzanïs, 1, pp 143-1 98. Ed by VC Mow and WC Hayes. New York, Raven Press, 199 1
Poole AR: Imbalances of anaboIisrn and catabolisrn of cartilage matrix components in osteoarthritis, In: Osreoarrhriric Disorders., V Kuettner, V Goldberg, and IL Rosemont. American Academy of Orthopedic Surgeons, 1995
Radin EL, Burr DB, Caterson B, Fyhrie D, Brown TD, Boyd RD: Mechanical deterrninants of osteoarthrosis. Seniin.Arthritis Rhezrm- 2 1 (3 Suppi 2):12-2 1, 199 1
S.J. Ferguson Queen's University Biomechanics of the Acetabular Labnrm
[30] RushfeIdt PD, Mann RW, Harris WH: lmproved techniques for measuring in vitro the geometry and pressure distribution in the human acetabulum - LI. instrumented endoprosthesis measurement of articular surface pressure distribution. Jozrrnul of Biomechanics 14:3 15-323, 198 1
[3 11 Strickland EM, Fares M, Krebs DE, Riley PO, Givens-Heiss DL, Hodge WA, Mann RW: In vivo acetabular contact pressures during rehabilitation, part 1: acute phase- Physical Tlrerapy 72:69 1-699, 1992
1321 Takechi H, Nagashima H, Ito S: Intra-articular pressure of the hip joint outside and inside the Iimbus. Journal of the Jnpanese Orthopaedic Association 56539-536, 1982
[32b] Tanabe H: Aging process of the acetabular labrurn - an electronic rnicroscopic study. Journal of the Japanese Orthopaedic Associarion 65:18-25, 199 1
Tepic S: Congrtlency of the hrrman hip joint, MSc, Thesis, Massachusetts institute of Technology. 1980
Tepic S: D_inamics of and entropy prodticrion in the carrilage [vers of rlle synovial joint., Sc-D. Thesis, Massachusetts institute of Technology. 1982
Terayama K, Takei T, Nakada K: Joint space of the human knee and hip joint under a static load. Engineering in Medicine 9:67-74, 1 9 80
von Eisenhan R Adam C,Steinlechner M, MüIler-Gerbl M, Eckstein F: Quantitative determination of joint incongruity and pressure distribution during simuiated gait and cartilage thickness in the human hip joint. Jorirnai of Orthopaedic Resenrch 17532-539, 1999
Weber W, Weber E: Uber die mechanik der menschlichten gehwerkzeuge nebst der beschreibung eines versuches uber das herausfailen des schenkelkopfes aus der pfanne im luflverdennten raum- Annalen Plysik und Chemie 40: 1-1 3, 1837
S.J, Ferguson Queen's University Chapter Eight
General Discussion
This thesis presents an investigation of labrum fùnction, a topic which has seen
only limited attention in the past. It is possible that the question of labnim fiinction has
simply been overshadowed by more obvious issues in research of the hip joint. While the
biomechanics of the hip joint have been extensively evaluated, the specific role that the
Iabrurn plays in the overall function of the joint has received almost no attention in the
literature. As most structures in the human body exhibit an elegant, and ofien complex
relationslip behveen forrn and fiinction, one may assume that it is this degree of subtlety
which accounts for the lack of knowledge of the acetabular labrum. The recent increased
interest in the consequences of acetabular labrum tears in the orthopaedic clinical
literature highlights this lack of a basic understanding of the fknction of the acetabular
labrum. Labrum tears are currently treated by partial labrum resection [17b], without a
&II understanding of the implications ofthis intervention.
A similar situation existed several decades ago in the treatment of meniscal tears.
The standard treatment then was total meniscectomy (meniscus resection). While the
short-terrn results were good, in the long term many patients developed premature
osteoarthrosis in the knee joint [2,5,20]. Only through careful study of the biomechanics of the knee joint was the role of the meniscus in load transfer and shock absorption in the joint discovered [1,7,9,12,18]. Biomechanics of the Acetabular Labrum
Drawing on a relatively sparse body of published work, several hypotheses were
Formulated for the possible function of the acetabular labrum in normal hip joint function,
and the influence of labnim pathology on the progression of the normal, healthy hip joint to the degenerative state of osteoarthrosis. It was hypothesised that: (1) the labrum seaIs a
pressurised fluid layer within the joint, preventins direct contact of cartihge surfaces, (2)
the labrum slows the rate of fluid expression fiom the cartilage layers, limiting overall cartilage deformation and stresses and (3)the labrum contributes to load transmission
within the hip joint.
The aim of this thesis was to follow a systernatic approach, using a combination of analytical rnodels and experimental methods, to evaluate these hdypotheses about
labrum function. The chapters of the thesis have been assembled in the order with which the individual studies were originally planned. Conceptually, there should be a logical progession tiom the results of one study into the goals and methods of the subsequent studies.
Anatomy Study
At the start of this project, there was no detailed anatomical study of the acetabdar labrum available in the literature for reference. Any descriptions of the labrum found in general anatomy texts were vague. Recently, a more detailed study of the labrum has been published by Putz et al 1161, but even this study leaves basic questions unanswered about the size, shape, and position of the acetabular labrum in the hip joint.
S.J. Ferguson Queen's University Biomechanics of the Acetabular Labmm
The study of Putz et uf [16] provided qualitative information about the general fom,
structure and composition of the labrum, but it lacked any detailed quantitative data.
The complex biomechanical function of the joint cmbe studied with the use of
computer models. For the generation of finite-element models of the acetabulum and
labrum, a technique would be required to accurately determine the overall three-
dimensional shape of the labrum, and its relationship to the other structures of the hip
joint. More conventional techniques for rapid segmentation of anatomical data were
considered, but these methods (CT. MRI) lacked the spatial resolution required for the
deveiopment of accurate rnodels, and did not provide adequate soft tissue visualisation.
With the emergence of the Visible Human project [14], a public-domain database of
high-resolution sedsection images of the hurnan body was made available, However,
even these images failed to provide adequate differentiation between the visually similar,
but structurally different soi3 tissues of the hip joint.
A technique which would allow tissue staining for contrast was required, and the
conventional technique of poIymer embedding was a logical choice. As there has been
some controversy in the Iiterature about the effects of polyrner embedding on the
dimensional stability of anatomicaI specimens, experiments were conducted to determine the level of tissue shrinkage one couid expect with specimens of mixed tissue type, such as whole joints. The overali level of tissue shrinkage / expansion was less than 2%. The choice of a three-part histological stain provided excellent soft tissue contrast. The technique was then applied to a whole human hip joint. This presented its own technical challenges, as polymer embedding of such a large specimen is not a common process, and therefore the initial attempts were time-consuming. However, the final sections
S.J. Ferguson Queen's University Biomechanics of the Acetabular Labrum
produced by this method were excellent. Reconstmction of the three-dimensional
c.geometry provided a basis for future solid models of the joint, and also basic dimensions
of the labrum for use in developing the simplified models of this thesis, The sections
showed that the labrum considerably increases the coverage of the femorai head by the
acetabulurn. The labrum, as observed in this study and in MRI scans which were also
examined, extends past the widest point of the femoral head (Le. equatorj and is tightly
apposed against the cartilage surface of the femur. The labrum has a substantial tissue
volume, greater than that of the adjoining acetabuIar cartiiage layer. The labrum is
triangular in cross-section, with a base width of approximatety 7mm and a heiglit of
lOmrn. For cornparison, the cross section of the meniscus has simiIar dimensions.
While the hip joint may be idealised as a simple ball-and-socket joint, its
eeometry is indeed quite complex (Figure 8.1). The spherical femorai head articulates Y against the horseshoe-shaped articular surface of the acetabulum; the depressed centre of the acetabulum (tossa) does not form part of the articular surface. A large portion of the
load applied to the joint is transferred across the superior region of the acetabulum. In this area, the labrurn forms a prominent extension of the articular surface, and could possibly contribute to load transfer wïthin the joint.
S.J. Ferguson Queen's University Biomechanics of the Acetabular Labm
A Acetabulum \,*A *' .
\. Transverse
Ligament \. \
Figure 8.1 : Anatomy of the hip joint, coronal section. While the hip rnay be considered a simple ball-and-socket joint, the articulation between the femur and acctabuium is rather cornplex. The labrum toms a signif-lcantextension of the acetabulurn and, consequently, rnay contribute to load transfer across the joint.
The labrum was of adequate size, and in the appropriate position, to fulfil the proposed sealing and load bearing functions, but anatomical data alone was not enough to
confirm the hypotheses. Knowledge of the inherent material properties of the labrum could provide further evidence of the proposed sealing and load bearing mechanisms, as well as providing the necessary data for detailed analytical models of the labrum, and its function in the hip joint.
S.J. Ferguson Queen's University Biomechanics of the Acetabular Labrum
Material Pro perties
A structure, whether it is engineered or biologïcaI, derives its strength and
stiffness from both its shape and its intrinsic materiai properties. There was evidence
fiom studies of the rnicroscopic structure of the labm [19] that it possessed a
specialised structure with some degree of anisotropy, similar to that of the
fibrocartilaginous meniscus. The predominant orientation of the major collagen fibre
bundles in the tissue was in the circumferential direction, around the rim of the
acetabutum. However, there was no published data on the material properties of the
labrum, or how these properties might relate to the anisotropy observed in its structure.
- For this thesis, a relevant subset of the complex rnaterial properties of the labrum were
determined. Aithough bovine specimens were selected for their availabiiity and size,
there is a wealth of published data on the properties of other bovine tissues and their
counterparts from human specirnens for comparison of the results.
Tensile properties determined from specimens o~entedin the circumferential
direction showed that labrum tissue has a stifiess and strength (75MPa and I2MPa
respectively) much higher than that of the adjoining cartilage iayers (approxirnately
SMPa and 7MPa respectively [17]). For comparison, the stiffness of the bovine meniscus
ranges fiorn 48 - 198MPa, depending on depth [15]. The ultimate strength of bovine
rneniscus ranges from 6 - 30MPa [ 151. The rneasured matenal properties of the labrum
were sipificantly greater (stifiess +44%, strength + 46%) in the posterior region of the
acetabulum than in the superior region. Based on the predominant loading direction in
S.J. Ferguson Queen's University Biomechanics of the Acetabular Labnun
the quadruped, it is possible that there is some spatial adaptation of labrum properties to
its mechanical environment.
The compressive properties of the labrurn were measured using a unique confined
compression apparatus which employed an ultrasonic displacement measurement to
minimise the influence of the testing apparatus on the consolidation experiments. The
compressive stifFness of the tissue, and its permeability to interstitial tluid flow, were
detennined in the direction of, and perpendicular to the predominant direction of fibre
orientation. A constitutive model, treating the tissue as a poroelastic solid, was used to
represent the fluid - solid interactions within the labrum. In compression. the labrurn has
an equilibrium modulus (O. l6MPa) approximately one third to one half that of articular
cartilage from similar specimens (0.29 - O.48MPa [3]). The permeability of the labrum
(5.0 x IO-'' rn"/~-s)was found to be much lower than that reported for the adjoining
articular cartilage layers (47 x 10-16 m4/N-s [3]). The labrum stifiess was lower than that
reported for the meniscus (O. 12 - 0.7MPa [8,15]) and the permeability of the labrum was
lower than that of the meniscus (8 - 35 x 10'16 m4/N-s [8,15]).
Labrum Function
From the anatomical study. it was observed that the labrum was largest in cross section in the superior region of the acetabulum. It would appear that the labmrn adapts
itself during development to the stimulus of the applied load across the joint. Labrum tissue may develop, increasing its size and perhaps collagen fibre density, thereby improving the intrinsic properties of the labnim, to better bear the loads which are
S.J. Ferguson Queen's University Biomechanics of the Acetabular Labm
applied across the joint. Through its substantial coverage of the femoral head, and a strong, fibre reinforced composition, the labrum may be able to resist dislocation of the femoral head from the acetabulum through the development of large hoop stresses within its ring structure, enhancing hip joint stability. Finite element analysis of this stability enhancing mechanism (Appendix B) has show, however, that the mechanical resistance to dislocation provided by the Iabrum itself is only a portion of the total resistance provided by the vacuum pressure which develops within the joint upon luxation.
Therefore, it would seem more important that the labrum, by means of its inherent elasticity, is able to maintain tight contact with the femoral head throughout the natural range of motion of the joint in order to provide this sealing function.
With a high resistance to interstitial fluid tlow, particularly in the radial direction, the labrum possesses the intrinsic properties required to form a relatively impermeable seal around the periphery of the acetabulum. Both its location in the joint and its cross- sectional shape are reminiscent of wiping seals in engineering applications. The Iabrum lias adequate cornpliance to follow the shape of the femoral head throughout its range of motion and yet may be stiFfenough in the circumkrential direction to maintain this seal.
The serial frozen sections of loaded hip joints presented by Terayama et al [22] support the premise that the labrum can seal the intra-articuIar space against the flow of pressurised fluid, and that the iabmm can maintain tight contact against the femoral head.
Indeed, Macirowski et al [l 11 have shown that, for contact of cartilage, after a short time the resistance to flow in the gap between the tcvo tissues is orders of magnitude higher than the resistance to tlow through the tissues themselves, so such a seal is plausible.
S.J. Ferguson Queen's University Biomechanics of the Acetabular Labrum
The anatorny and rnaterial property studies have demonstrated that the labmm is a unique structure within the hip joint, with a forrn and inherent properties which irnpIy a
fùnction consistent with the hypotheses of this thesis. On the one hand, the labmm provides a substantial estention of the acetabular nm, and under certain loading conditions couId play a role in the transfer of load across the joint. Perhaps more importantly, the labmm, by virtue of its low-permeabilty, may present a barrier to tluid flowing fiom within the joint. With compelling evidence that hip joint cartilase layer consolidation, and there fore cartilage tissue de formation and stress, is determined b y the resistance to interstitial fluid flow out of the cartilage layers [ 1 11, the influence of the labrum on this fluid flow process became the focus of subsequent cornputer and experimental analyses.
Finite Element Models
The hypotheses on labrum hnction have been explored using cornputer models.
An cLvisyrnrnetnc mode1 of the hip joint cartilage layers and labrum was devdoped to determine whether or not the labrum could seal a pressurised layer of fluid within the joint space, as one could infer from the work of Terayama et al [22]. This sealing mechanism is shown schematically in Figure 8.2.
S.J. Ferguson Queen's University Biomechanics of the Acetabular Labm
Figure 8.2: The labmm acts as a "wiping seal", slowing or completely preventing flow of synovial fluid from beîween the joint surfàces. Maintenance of this pressurised (Po) fluid layer may limit stress levels within the cartilage layers.
The results of the mode1 showed that such a seal could exist, but that this mechanism would be dependent on the circum ferential sti fiess of the labrum fibres. The measured values of tensile sti ffness of the bovine labnirn specirnens showed a large variation.
Many specimens had a measured stiffness in the range for which the mode1 predicts a sealinz function could exist. Hip joint geometry would also be a determining factor for
Sf- Ferguson Queen's University Biomechanics of the Acetabular Labmrn the effectiveness of this sealing mechanism; the sIight incongruity of the hip joint wouId enhance such a seal. Over time, this seal is gradualiy lost, as tissue creep and fluid redistribution overcome the contact pressure holding the Iabrum against the tèmur. The model predicted that. in the presence of such a sealing function, loads would be transferred across the joint through a pressurised fluid layer, resulting in low surface and subsurface stresses in the cartilage layers themselves. While this model was useful for testing the basic hypothesis of Iabrum sealing, the limitations of the finite eIement software did not allow subtle aspects of such a sealing mechanism to be studied. For example, the mode1 could only predict the presence or absence of such a seal, but a failure of this sealing mechanism would iikely be gradual, with some leakage of fluid out of the joint space. What would the consequences be of a "leaky" seal? The model could not predict the intluence of the labrum on thin lubricating tluid films of the same scale as the surface irregularities of the cartilage layers themselves.
Independent of the ability of the labrum to seal a lubricating layer between the tu.0 surtàces of the hip joint, the Iabrum could possibly influence the overall rate of cartilage layer compression within the hip joint, in response to applied loads.
Macirowski et al [Il] showed that, once two opposing cartilage surfaces corne into contact. the expression of interstitial fluid fiom the cartilage. and hence the overall deformation of the tissue, is govemed by the resistance to flow within the tissue layers.
The resistance to flow through the gap between the two cartilage layers is much higher than the resistance to flow through the layers themselves, and therefore the predominant flow path is tangentially, through the tissue. A model was developed to test the hypothesis that the labrurn, by means of its position, shape and low permeability,
S J-Ferguson Queen's University Biomechanics of the Acetabular Labrum
provides an additional resistance in this flow path, limiting the rate of cartilage layer
de formation,
This mechanism for limiting cartilage consolidation is shown schematically in
Figure 8.3. The results of the mode1 showed that, for loads approximating the residual
load applied to the joint throughout the day, the rate of cartilage layer consolidation was
up to 40% quicker followirg labrum excision.
Figure 8.3: The Iabm provides an additional resistance to fluid expression fiom the cartilage Iayers, by increasing the rate of the low-permeability flow path. This may result in a slower rate of cartilage layer consolidation, and consequently lower levels of tissue deformation and stress.
SJ . Ferguson Queen's University Biomechanics of the Acetabular Labrum
A portion of the load applied to the joint \vas camed by the labmm, as demonstrated by the distribution of contact pressures across the joint. Removal ofthe labrum resulted, as one would expect, in a redistribution of this load onto the remaining cartilage layer. The
Iocal solid contact stresses within the joint were up to 92% higher in the absence of the labrum. As the excellent fiictional properties of cartilage are determined by the normally low level of contact stress between the solid coIlagenous matrix components at the tissue surfaces [13], increased solid contact stresses could initiate the damage which has been observed in conjunction with labrum patholog through increased frictional Wear. Strains and stresses within the cartilage Iayers were higher without the labrum, a possible factor in the initiation of subsurface fatigue damage of the collagenous matrix. indeed, elevated shear stress leveis at the labrum / cartiIage / bone interface correspond to the clinically observed delamination of cartilage at the acetabular rim.
To the author's knowledge, there have been no previous finite element models of the hip joint which have considered the fluid / solid nature of the soft tissues, and the developing contact and interaction of two natural cartilage layers. However, the elegant models of a single acetabular cartilage layer, developed by Macirowski [IO] and Tepic
[2 11, must be acknowledged. These were pioneering works, which introduced this author to the concept of "sealing" in a diarthrodial joint. More recent finite element modeis of joint contact biomechanics have also presented simplified, two-dimensional geometries
[4,23,24], highlighting the limitations of current sobvare and hardware for modelling the interaction of two three-dimensional poroelastic joint surfaces. The rnodels developed in this thesis have attempted to advance the use of commercial software by modelling the interaction between the joint surfaces and an entrained fluid layer.
S.J. Fereauson Queen's University Biomechanics of the Acetabular Labnim
Ideally, the hvo models in this thesis should be combined into one three- dimensional model, which couid accurately predict the influence of the labrum on the initial pressurisation of the intra-articular space, the eventual depletion of this fluid layer, the subsequent developing contact area benveen the joint surfaces and the long-term consolidation of the joint cartilage Iayers, Such a miael is not yet possible with commercially available sofbvare.
In Vitro Consolidation Experiments
The results of the finite element models have shown the importance of fluid pressurisation in the transfer of loads across the hip joint, and the role that the labmm plays - efficient sealing of possible fluid escape routes. As discussed, hvo separate models were used to evaluate this function of the labrum. Ideally, a single mode1 should be developed to observe this sealing rnechanism over tirne. While that was not possible, the results of the finite element analysis provided the motivation to develop an it~vitro experimental model to observe the influence of the labrum on overall hip joint biornechanics. This was the first experimental evaluation of joint consolidation which examined the influence of an accessory structure of the joint on the overall hip joint biomechanics. This esperiment was also the first to study fluid pressurisation within the joint and its relationship to cartilage layer consolidation.
The ovenvhelrning impression that one has of the results of these experiments is that the human hip joint is a remarkably well adapted structure. After an hour of continuous loading, the cartilage Iayer consolidation had not reached equilibrium. With
S.J. Ferguson Queen's University Biomechanics of the Acetabular Labrum consolidation displacements consistent with those predicted by the computer models, it could be assumed that the majority of the load transferred across the joint was carried by tluid pressure, rather than by stresses in the solid matrix of the cartilage layers, even after this extremely long Ioading period- The resuIts of this expenment provided independent confirmation of the hypotheses about labrum fùnction, and were consistent with the predictions of the models. It was show that, in the natural human hip joint, the initial consolidation displacernent rate \vas approxirnately 22% quicker following Iabnim excision. Pressure measurements taken during consoiidation experiments provided some evidence that the intra-articular fluid of the hip joint can be pressurised upon initial load application. However, these resuits depended on the fit bebveen the femoral head and the acetabular labrum. A correspondence was found between the increase in the intra- articular pressure decay rate and the increase in the initial joint consolidation displacement rate following labrum excision.
The finite etement models predicted two possible mechanisms by which the tabrum could influence the rate O t' cartilage layer consolidation in the hip joint: by sealing a pressurised fluid layer within the joint, preventing fluid flow pust the labrum, for a duration on the order of minutes, and by decreasing the long-term rate of cartilage Iayer consolidation through additional resistance to interstitial fluid flow throzrgh the tissue, for a duration on the order of hours (Figure 8.4)- Significant changes in the esperimentally observed joint consolidation rate occurred in the short-term, and were accompanied by changes in the fluid pressurisation decay rate. Long-term changes in joint consolidation and fluid pressurisation observed in the experiment were not significant, From these
S.J. Ferguson Queen's University Biornechanics of the Acetabular Labrum
observations. one can infer that the dominant mechanism by which the labmrn influences
overall hip joint biomechanics is in the initial sealing of the intra-articular space.
However, as discussed previously, the influence of the labnirn on long-term cartilage
layer consolidation is dependent on joint loading direction. For certain loading conditions, the labnirn may have a more pronounced influence on cartilage layer
consolidation.
time
Figure 8.4: Overall joint consolidation, as measured experimentally, rnay be govemed by two distinct mechanisms. The short tem, fast consotidation response of the joint rnay be determined by the rate of fluid expression fkom between the joint surf&s. The labim seals fluid within the joint space and limits the rate of this fluid layer's depletion. The long tem, slow consolidation response of the joint rnay be detemined to the resistance to flow of fluid being expressed from within the cartilage layers themselves. The labrum adds an additional resistance to the flow path of interstitial fluid. Experiments demonstrated that the consolidation of the joint was faster following tabmm tesection.
SJ . Fer,gxon Queen's University Biornechanics of the Acetabular Labrum
Clinical Significance
What are the implications of these findings? Through the work of this thesis, it has been established that the labrum is a substantial component of the hip joint, with a unique structure and properties presumably tailored to a specific function(s).
Experimental and analytical evaluation of labrum function has provided reasonable evidence for the proposed biornechanical rote of the labrum. The labnim can fonn a sea1 around the fernord head, effectively blocking fluid escape routes f?om the joint, enhancing the protective mechanism of fluid pressurisation within the hip joint cartilage layers. The labrum also carries a portion of the load applied to the hip joint.
How the failure of this function could lead to degenerative changes in the hip joint, consistent with clinically observed damage, remains controversial. It is plausible that the increased Ievels of cartilage consolidation associated with labmm pathology could eventually lead, through increased levels of stress within the tissue, to cartilage fibrillation and delamination. However, the possibiIity exists that degeneration of the cartilage layers of the hip joint is a result of a number of other factors, such as joint geornetry or joint loading, independent of the labrum's contribution to overall joint function. Labrum pathology may even be an end result, and not a cause, of the degenerative changes within the hip joint,
Subjective evaluation of the experimental observations suggested that this sealing function is not present in al1 hip joints, but rather is dependent on the quality of the fit between the labrum and the femoral head. The current treatment of labrum tears is by partial or total labrum resection. While successful in the short term, it is possible that, by
S.J. Ferguson Queen's University Biomechanics of the Acetabular Labmm
compromising the sealing function of the labrum, this treatment accelerates the joint
degeneration that one had hoped to alleviate. With continued study of the function and
importance of the labnirn, new surgical repair stratepies rnay be developed to maintain
the overall function of the hip joint. It may be possible, through arthroscopie surgery, to
improve the fit of a '-loose" labrum to restore its sealing function. In the case of partial
labrum resection, it rnay be possible to replace missing tissue with an in vitro engineered
tissue, or with a biocompatible hydrogei. Such materials exist, with compressive
properties similar to cartilage and fibrocartilaginous tissues [6],and perhaps it is only a
matter of time before these materials develop the tensile stiffness and strength of labrum
tissue.
The findings of thiç thesis may also provide inspiration for improvernents in the
design of current hip prostheses. A labrum is not necessarily required in a conventional prosthesis, although there is evidence that the labrum is beneficial in the specific case of a
femoral endoprosthesis, as the overalI fùnction of an artificial joint is completely different than that of the natural joint. However, consideration of the results of this thesis may provide ideas for a new concept of prosthesis design. Wear of the prosthesis articulating surfaces is the first step in a process which eventually leads to implant failure through component Ioosening. Prosthesis wear. resulting fiom the same mechanisms of fnctional adhesion and subsurface delamination discussed previously for cartilage, is determined by the stresses at and below the contact surfaces. A prosthesis, which incorporates a peripheral seal between the acetabular and fernoral component, could benefit from the advantages of a pressurised fluid Iayer for load transmission, which have been discussed in this thesis in the context of the natural joint.
S.J. Ferguson Queen's University Biornechanics of the Acetabular Labrum
References
Ahmed AM, Burke DL: in-vitro measurement of static pressure distribution in synovial joints - Part 1. Tibia1 surface of the knee. Jozrrnal of Biomechanical Engineering 105:2 16- 225,1983
Allen PR, Denham RA, Swan AV: Late degrnerative changes afier meniscectomy. Factors affecting the knee afier operation. Journal of Bone and Joint Szirgery [Br] 66-B:666-67 1, 1984
Athanasiou KA, Agarwal A, Muffoletto A, Dzida FJ, Constantinides G, Clem M: Biomechanical properties of hip cartilage in experimental animal models, Ciinical Orthopaedics and Related Research 3 16:254-266, 1995
Donzelli PS, Eckstein F, htz R, SpiIker RL: Physiological joint inconpity significantly affects the load partitioning between the solid and fluid phases of articular cartilage. Proceedings of the 43rd Annual Meeting of the Orthopaedic Research Society, San Francisco, California, pp. 82, 1997
Fairbank TJ: Knee joint changes after meniscectomy. Jozrrnai of Bone andJoinr Srrrgery [Br] 30-B:664-670, 1948
Ferguson, S- J- and Ito, K, Mechanical Properties of Hydrogels. A0 Report, 1997
Fukubayashi T. Kurosawa H: The contact area and pressure distribution pattern of the knee. A study of normal and osteoarthrotic knee joints. Acta Orrhopaedica Scandinavica 5 1:87 1-879, 1980
Joshi MD, Suh J-K, Marui T, Woo SL-Y: lnterspecies variation of compressive b iomechanical properties of the meniscus. Journal of Biomedical fifaterials Research 29:823-828, 1995
Kurosawa H, Fukubayashi T, Nakajima H: load-beanng mode of the knee joint: physical behaviour of the knee joint with or without menisci, CZinicai Orthopaedics and Relured Research 149:283-290, 1980
Macirowski T: Stress in the cartilage of the hzrrnan hip joinr, D.Sc. Thesis, Massachusetts Institute of Technology. 1983
Macirowski T, Tepic S, Mann RW: Cartilage stresses in the human hip joint. Jortrnal of Biomechanical Engineering 1 1 6: 1 0- 1 8, 1 994
Maquet PG, Van de Berg AJ, Simonet J: Femorotibial weight-bearïng areas. Experimentd determination. Journal of Bone andJoint Srrrgery 57-A:766-772, 1975
McCutchen CW: The tiictional properties of animal joints. Wear 5: 1-1 7, 1962
NLM. The Visible Human Project, National Library of Medicine . 1999. Ref Type: Electronic Citation
S.J. Ferguson Queen's University Biomechanics of the AcetabuIar Labmrn
[15] Proctor CS, Schmidt MB, Whipple RR, Kelly MA, Mow VC: Material properties of the normal media1 bovine meniscus. Journal of Orthopaedic Research 7:77 1-782, 1989
[16] Putz R, Schrank C: [Anatomy of the labrocapsular complex of the hip joint]. Orthopcïde 27:67S-680, 1998
[17] Roth V, Mow VC: The intrinsic tensile behavior of the matrix of bovine articular cartilage and its variation with age, Journal of Bone andJoinr Szrrgety 624:1 102- 1 1 1 7, 1980
[17b] Santon N, VilIar RN: Acetabular labrum tears: results of arthroscopic partial Iimbectomy. Arthroscupy 16: 1 1-1 5,2000.
[18] Seedhom BB: Transmission of the load in the knee joint with special reference to the role of the menisci, Part II: experimenta1 results, discussion and conclusions. Engineering in Medicine 8220-228, 1979
[19] Shibutani N: Three-dimensional architecture of the acetabular labrum - a scanning electron microscopic study. Journal cf the Japanese Orrhopaedic Associaiion 6233 1-3 29, 1 988
[20] Tapper EM, Hoover NW: Late results after meniscectomy. Jotrrnal of Bone and Joinr Szrrgety 5 1-A:5 17-526, 1969
[2 1] Tepic S: Dynarnics of and entrop-v production in the cartilage luyers of the synoviu[joint., Sc-D. Thesis, Massachusetts lnstitute of Technology. 1982
[22] Terayama K, Takei T, Nakada K: Joint space of the human knee and hip joint under a static Ioad. Engineering in Medicine 9:67-74, 1980
[23] van der Voet AF, Shrive NG, Schachar NS: Numerical modelling of articular cartilage in synovial joints - poroelasticity and boundary conditions. In: Cornptïrer Merhods in Biomechclnics and Bionredical Engineering, 200-209. Ed by J Middleton, GN Pande, and KR Williams. Swansea, IBJ Publishers, 1993
[24] WUJZ, Herzog W, Epstein M: Evaluation of the finite eIement sofhvare ABAQUS for biomechanical modelling of biphasic tissues. Journal of Biornechanics 3 1 :165-169, 1998
S.J. Ferguson Queen's University Chapter Nine
Conclusion
Summary
The objective of this thesis was to combine experimental and analytical techniques to study the function of the acetabular labrum in the normal hip joint. and
f?om this to infer the possible role that labrurn padiology could play in the initiation and progression of osteoarthrosis. When this thesis research was proposed, there \vas little known about the labnirn, its structure and composition, its mechanical properties, or its influence on hip joint biomechanics. In the foIlowing years, little has changed in this regard- However, the labrum has received increased attention in clinical orthopaedic literature. Tears of the labrum have been cited as a cause of hip pain, and as a possible precursor to osteoarthrosis in the hip joint. The operative treatrnent of labral tears consists primarily of partial or total resection. However, while the short-tenn results have been good, the long-term effècts are not known.
Based on the c!inical evidence of a possible link between labrum pathology and cartilage degeneration, and the limited experirnental investigation of labrum function reported to date, several hypotheses for the function of the acetabular labrum were proposed. The main hypotheses of this thesis were that: (1) the Iabmm seaIs a pressurised fluid layer within the joint, preventing direct contact of cartilage surfaces, (2) the labrum slows the rate of fluid expression fiom the cartilage layers, Iimiting overall Biomechanics of the AcetabuIar Labrum
cartilage deformation and stresses and (3) the labrum possesses adequate structural
properties to cany a portion of the load applied to the joint.
The first problem faced when studying the function of the labrum was the
complete lack of detailed quantitative data of labrum anatorny. The first section of the
thesis concems the deveIopment of a histological technique for prepanng and sectioninz
large anatomical samples for reconstruction of their three-dimensional anatomy. This technique \vas applied to a whole human hip joint. The Iabrum has a substantial tissue
volume and was found to considerably extend the coverage of the acetabulum over the
femoral head.
The second section of the thesis presents an investigation of the tensile and compressive material properties oflabrum tissue, as there have been, to date, no such published studies. The results of this investigation provided some insight into the nature oflabrum tissue, and also provided parameters for the analysis of labrum function. The results of these experiments demonstrated that the labrum, with its highly oriented collagen fibre structure, was much stiffer and stronger than the adjoining cartilage. The resistance to fluid flow through the labrurn was also much higher than through cartilage.
One can in fer that the strength and impermeability of the labrum enhance its ability to seal and stabilise the hip joint.
However, knowledge of the inherent mechanical properties of the labrum alone was not adequate to evaluate our hypotheses. Poroelastic finite-element models of the hip joint were deveIoped to study the influence of the labrum, as a structure, on overall joint biomechanics. The models dernonstrated that the labrum can seal a pressurised fluid layer within the hip joint under loading for a physiologically relevant period of tirne.
S.J. Fer,pson Queen's University Biomechanics of the Acetabular Labrum
Consequently, cartilage stresses and contact pressures were reduced. The rnodels also
demonstrated that, with its low permeability, the labrum adds an important resistance to
the flow path for fluid expressed from the cartilage layers. Stresses and strains within the
cartilage layers calculated by the mode1 were up to 4 1% higher following removal of the
labrum. Contact pressures, and hence fkiction between the cartilage surfaces, were also
significantl y higher foollowing labrum removal.
In order to further explore the predictions of the finite element analyses, a series
of in vitro whole-joint creep consolidation experiments on human hips before and after
labrum excision were conducted. The overall compression of the cartilage layers under a
variety of static and dynarnic loads was measured, as was the fluid pressure in the fossa
-of the joint, and by extension the intra-articular fluid pressure. The experiments
demonstrated that removal of the labrum resulted in a quicker cartilage consolidation
rate. Greater deformation of the cartilage layers implies higher stresses within the tissue.
Peak Eluid pressures of over 500 kPa were measured during loading in joints with a well-
forrned labrum. Chslnges in the rate of pressure decay before and after labrurn excision
corresponded to changes in the initial consolidation displacement rate, implying that
initial sealing of the intra-articular space by the labrum is the dominant mechanism
through which overall joint consolidation was regulated. The results of the experirnents
were consistent with the predictions of the finite element models and lend further support
to Our hypotheses about labrum sealing.
The aim of this thesis was to follow a systematic approach to evaluating several hypotheses about labrum function by first determining the basic material properties of the
labrum, then applying this knowledge to analyticaI models of labrum function, and finally
S.J. Ferguson Queen's University Biomechanics of the Acetabular Labrurn
conducting a series of experiments to validate the analytical results. With an understanding of labrum function, guidelines may be suggested for surgical intervention in the case of labrum pathology.
Future Work
The work presented in this thesis was only a step along the way to a better understanding of the function ofXe acetabular labrum, and the complexities of hip joint biomechanics in general. The following studies are proposed to extend the work of this thesis and to broaden the scope of füture investigations into labrurn function:
1. Further characterisation of the material properties of the labrum. In this thesis,
a specific set of the materia1 properties of the bovine acetabular labrum were studied.
If adequate testing equipment could be developed for the relatively smaller
specimens, human labrum tissue should be measured. Compression testing should be
extended to determine whether or not the properties of the labrum tissue are strain
dependent, as has been observed with cartilage specimens. Indentation tests or
optical measurements should be conducted to determine the Poisson's ratio of labrum
tissue. Tensile testing should be extended to determine the effects of strain rate on
the measured properties of iabrum tissue, and also to detemine its time dependent
viscoelastic tensile properties-
SJ. Feraouson Queen's University Biomechanics of the Acetabular Labm
2. Further experimentat measurement of whole-joint consolidation. Human hip
joint specimens can be difficult to obtain, hence the limited sample size (n = 6)of the
consolidation experiments in this thesis. These experirnents should be extended to
include additional specimens. The experiment itself could be improved through
better placement of the pressure transducer, by using a radiographic image-
enhancement technique during placement to ensure that the transducer is inserted into
the fossa. A method for quantiwng the "fit" of the labrurn on the femoral head
should be deveIoped. Pre-testing MRI and CT scans of the joint would provide the
necessary information to investigate the possible link between joint con-mency and
joint sealing. Finally, the method proposed in Appendix E for directly measuring the
consolidation of the cartilage layers should be used if the expected improvernents to
the QuaIisys system are realised.
3. Three-dimensional poroelastic model of labrum function. The fmite element
models presented in Chapters Five and Six were simplified, two-dimensional
representations of the hip joint cartilage Iayers and labrum. Limitations on computing
power and on the available interactions between poroelastic eiements alon,a contact
surfaces within the ABAQUSsoftware made such simplifications necessary. With
current increases in cornputer storage and processing power, together with
improvements in the non-Iinear contact algorithms of commercial finite element
software, extension of these simplified rnodels to a full three dimensional model of
the hip joint shouId be feasible. Using the quantitative data of the human hip joint
(Appendix A), such a model, which would include the bony structures, cartilage
layers and the labrum, could be constmcted. With this model, and enough solution
S-J. Fermouson Queen's University Biomechanics of the Acetabular Labrum
tirne, the hip could be subjected to a typical day's loading, with varying load direction
and magnitude, to better evaluate the hypotheses of this thesis about Ion,-O term
cartilage layer consolidation in the hip joint, with and without the labrurn.
4. Tribological analysis. Earlier analyses of the possible lubrication regimes in the hip
joint treated the joint as a simple, congrnent bal1 and socket joint. Inclusion of the
labrum structure in a mathematical model of hip joint lubrication, with appropriate
geometry, roughness and cornpliance of the two opposing joint surfaces, may provide
new insight into the mechanisms responsible for the low fiction and long liîè of the
human hip joint.
5. Analytical and experimental investigation of joint stability. It has been proposed
that the labrum enhances joint stabiIity and can help to prevent joint luxation. Using
the three-dimensional model proposed in (3), the mechanical contribution of the
labrum to joint stability could be evaluated for a variety ofjoint motions. Joint
motions which cause luxation are short-duration events, therefore the tirne-dependent
poroelastic properties of the cartilage and labrum would not have to be considered in
such a model; instead, appropnate elastic material properties couId be substituted.
T'ensile properties for the quasi-static tests of Chapter Four would provide a
conservative estimate of the tensile stifiess and strength of the labrum at high strain
rates. Equivalent elastic compressive properties of the labrum could be estimated
from a simple poroelastic model of a tissue specimen loaded at a high main rate.
Depending on the results of computer simulations, a full series of experiments using
cadaver joints could be designed which would include effects which the computer
S.J. Ferguson Queen's University Biomechanics of the Acetabular Labm
models may not have the ability to simulate, such as the development of a negative
pressure in the acetabulum during joint luxation.
6. Finite element model of labrum injury. It would be oFclinical interest to
understand the mechanism of Iabnim injury (tears, hypertrophy). Using a simplified
three-dimensional model of the joint, a variety ofjoint geometries, motions and Ioads
could be evaluated. A labrum tear rnay be the result of a single excessive load
appIied to the joint, it rnay result fiom the joint being moved to an extreme position,
or it rnay be the result of cumulative fatigue damage fiom impingement of the
femoral neck on the labrum. Currently, some clinicians are performing corrective
surgery to re-shape abnonnally thick femoral necks, or to realign the acetabuium or
proximal Femur to prevent just such an impingement. However, these surgeries are
performed with no clear evidence of the causes of labrum injury.
7. Animal mode1 of osteoarthrosis induced by labrum injury. Moving beyond a
purely engineering study, an in vivo animal model could be used to evâluate the
hypothesis that iabrum injury is a direct cause of osteoarthrosis, as has been implied
in the clinical studies referenced throughout this thesis. With knowledge of a
mechanism for labrum injury gained through the analyses of (6), an in vivo model
could be deveIoped whereby Iabrum injuries could be induced in an experimentai
model, and subsequent changes to the cartilage layers of the hip joint could be
studied.
S.J. Ferguson Queen's University Appendix A: Quantitative Anatomical Study
of the Human Hip including the Labrum
The histological preparation and 3D reconstruction techniques described in
Chapter Three were applied to a human hip specimen to determine the shape, size and position of the acetabular Iabnim in the human hip joint. From the whole pelves obtained for the consolidation experïrnents of Chapter Seven, the left hip of a 78kg male, aged forty at the time of death, was selected for the quantitative anatomy study. The specimen was obtained within twenty-four hours post-rnortem and was stored at -20°C pnor to preparation. Prior to embedding the joint was thawed and cleaned of ail soft tissues, except for the joint capsule and labrum. A 2mm Kirschner wire was driven into the inkrior anterior spine, aligned in the posterior-anterior direction (perpendicular to the coronal plane). This Kirschner wire would serve as a reference for later alignment of the reconstructed geometry. Excess bone was cut from the fernur and pelvis, with saw cuts made distal to the greater trochanter, and through the ilium, pubis and ischium. The subsequent fixation and embedding steps were performed as outlined in Chapter Three.
Polymerisation was initiated at 5°C and completed under vacuum at room temperature to prevent the formation of large air bubbies.
Following polyrnensation, the joint was sectioned on an precision band saw with diamond blade (Exakt 3 lOCP, Exakt Apparatebau GmbH, Norderstedt, Gerrnany) at intervais of 1.5 mm (Figure A. 1). The sections were then polished on a water-cooled, fine grit rotary ginding table (Struers, RotoPol-25, Copenhagen, Denmark) and stained with the three-part surface staining procedure described in Chapter Three. This stain Biomechanics of the Acetabular Labm provides excellent contrast between sofi-tissue types which are cannot be visually differentiated in the &esh specimen, for example the labrum and cartilage (Figure A.2)
Figure A. 1: The polymer embedded hip specimen was sectioned at 1.5mm intervals using a precision ban saw with diamond blade.
Following staining, individual sections were placed in an alignrnent heand an image taken with a CCD digital colour camera (Minolta RD- L 75, Minolta (Schweiz) AG,
Dietikon. Switzerland), 50mm macro lens, with a 24-bit colour depth and at a resolution of 1600 x 1200 pixeIs. Using the SuRFdriver software [!], contours were manually superimposed over the sofi-tissue structures - cartilage and labrum - and the underlying
S.J- Ferguson Queen's University Biomechanics of the AcetabuIar Labm
bone. foflowing visible lines of colour and tissue morphology separation. These contours
were then exported as IGES data for importation into solid modelling software. The
SuRFdriver software was used to make quantitative measurements of total tissue volume and contour areas, while labrum height and width and cartilage layer thickness were measured directIy from the stained sections.
Figure A.2: Stained section throu* the coronal plane of the hip joint. The bone (B), labnim (L) and articular cartilage (AC) are stained different colours according to their tissue type.
S.J. Feraauson Queen's University Biomechanics of the Acetabular Labrum
Measurements taken ftom the individual sections showed that the acetabular labrum substantialIy increases the coverage of the femoral head by the acetabutum. The femoral head had a diameter of S2rnm. In the superior portion of the acetabulum, the labrum extended IOmm frorn the osseous margin of the acetabular socket, coveting an additional 22" sector of the fémoral head. The labrum in this region had a base width of
7mm, and the fibrocartilage tissue of the labrum extended in fkom the bony acetabular rim to blend with the acetabuiar cartilage layer. Maximum cartilage layer thickness was
3.0mm on the femur and 2.8mm in the acetabulum, The labrurn, together with the transverse acetabular ligament, wrapped beyond the equator of the femoral head (Figure
A.3).
Figure A.3: Mid-coronal section of the hip joint showing the extended coverage of the labmm (L) over the femoral cartilage layer (FC). The labrum blends wiùi the acetabular cartilage (AC). Together with the transverse acetabutar ligament (TAL), the Iabrurn wraps around the femoral head beyond its equator.
S.J. Ferguson Queen's University Biomechanics of the Acetabular Labnim
The acetabular labrum has a substantial tissue volume, as calculated by the
SURFdriver software. The total volume of the labrum and transverse acetabular li,aament was approximately 6760 mm3. The total volume of the acetabular cartilage layers was approximately 5020 mm3, and the volume of the femoral cartilage layet was 8050 d.
The labmm's position and tissue volume may enable it to provide a shock absorbing function within the hip joint. The contours extracted from the sections using the
SURFdriver software are shown in Figure A.4. These contours will be used in the future to construct a three-dimensionaI poroelastic finite element mode1 of the hip joint.
A.4: Boundary contours extracted from the serial sections of the hip joint. The labrum and the bony surfaces ofthe acetabulum and femoral head are shown, displaced for clarity. The acetabular and Ièmoral cartilage Iayers are not shown in this image.
References
[l] Moody, D. and Lozanoff, S. SURFdriver sofrware, 1999 (h~://www.surfdriver.com)
S.J. Fereauson Queen's University Biomechanics of the Acetabular Labrum
Appendix B: Analysis of the Mechanical Stability
of the Acetabular Labrum
In Chapter Four, a study of the material properties of the acetabular labmm kvas presented. In subsequent chapters, the influence of the low permeability and hi& circumferential stiffness of the Iabmm on sealing of the hip joint has been demonstrated.
What other purposes could be served by the acetabular labrum, with a structure unique to that of the adjoining cartilage layers? To investigate the potential of the acetabular
labrum to stabilise the femoral head in the acetabular socket, a simpIe finite elernent analysis was conducted.
A linear elastic, axisymmetric model of the femoral cartilage surtace, acetabular cartilage surface and acetabular labrum was created. The fernoral head diameter kvas
46mm and cartilage layer thickness was 3 mm. The Iabrum was modelled as a transversely isotropic matenal, with a circumferential stiffness of 70 MPa and a stiffness of 10 MPa perpendicular to the circumference. The cartilage layers were modelled with an effective elastic stifiess of 10 MPa. No effect of strain rate on material properties was modelled. The contact surfaces between the femur and acetabulum were modelled by contact elements with large sliding capabilities. The contact was assumed to be fictionless. The fèmoral head \vas dispIaced dong the model axis of symmetry, and the
Ioad versus displacement curve was recorded. The displacement of the model is shown in Figure B. 1.
S.J. Ferguson Queen's University Biomechanics of the Acetabular Labrum
Figure B. 1 : The axisymmetric mesh of the mode1 included a portion of the femoral and acetabular cartilage layes and the triangular acetabular labrum. The mode1 mesh is rotated 360" about the joint axis of symmetry for the soIution, Model displacements are shown as the femur is pulIed out through the acetabular labrum.
The force-displacement curve calculated for the finite eIement analysis is shown in Figure B.2. The force required to displace the femoral head from the acetabular socket increases up to a displacement of approximately 2mm, where the peak force is approximately 18N. Deformation of the labrum as the femoral head is pulIed through results in a hoop stress being developed in the predominantly circumferential fibres of the labrum, and a contact pressure between the labrum and femur. Due to the shape and position of the labrum. the vertical component of the resultant force acting on the femur resists the displacement of the femoral head. Beyond 2mm displacement, as the widest portion of the femoral head moves through and beyond the fernur, this resisting force decreases. If the femoral head is displaced Far enough, the resultant force eventua1Iy passes through zero and the femoral head is pushed out of the socket.
S.J. Ferguson Queen's University Biomechanics of the Acetabular Labrum
Fernoral Head Pul [-Out Force
15 -
t- I g 'O- P *
5 -
displaœment (mm)
Figure B.2: Force required to pull the fernoral head past-theacetabular labrum, out of the acetabular socket. A peak force of approximately 18N is reached at 22mm of displacement.
The studies of Takechi et d[2]and Weber and Weber [3] demonstrated that the labnim conrnbuted to the stability of the hip joint through the development of a negative pressure within the joint upon luxation of the femoral head. More recently, a detailed series of rneasurements have been made of the negative pressure deveIoped within the hip joint upon joint iusation [l]. Consistent with our observations during joint positioning
(see Chapter Seven), negative pressures of up to 0.8 bar were measured during controlled luxation, on a MTS testing fiame, of the femoral head fiom the acetabular socket. The pull-out force in these experiments reached its peak value for displacements in the range of 2 - 3 mm, An estimate of the contribution of the labrum's mechanical resistance to luxation could be made by comparing the total measured pull-out force to the measured vacuum pressure, integrated over the projected area of the femoral head. It was estimated
S.J. Feraauson Queen's University Biomechanics of the Acetabular Labrum
that the mechanical contribution of the labnim's mechanicai stiflhess to the total pull-out force was approxirnately 15 - 25 N, consistent with the predictions of the model. This represented approximately 10 - 15% of the total pull-out force, a srnall but appreciable contribution. However, it would seem that the labrum7smore significant contribution to joint stability would be its ability to conform to and sed against the femorai head, facilitating the development of a vacuum pressure within the joint upon luxation of the femorai head-
References
[l] Speirs, AJ, Müller Institute FUr Biomechanik: Personal Communication, 1998.
[2] Takechi H, Nagashima H, Ito S: Intra-articular pressure of the hip joint outside and inside the limbus. Jozrrnal of the Japanese Orthpaedic Association 56529-536, 1982
[3] Weber W, Weber E: Uber die mechanik der rnenschlichten gehwerkzeuge nebst der beschreibung eines versuches uber das herausfallen des schenkelkopfes aus der pfanne im Iuftverdennten raum. Annalen Physik und Chernie 40:l-13, 1837
S.J. Ferguson Queen's University Biomechanics of the Acetabular Labrum
Appendix C: Evaluation of ABAQUS Finite-Element Software for
the Solution of Problems involving "Biphasic" Soft Tissues
The ability of ABAQUS commercial finite element (FE) software to solve problems in biphasic poroelasticity is examined. Mow et ûI [l] developed their biphasic theory and have applied it to the analysis of cartilage and load bearing sot? tissues. The poroelasticity theory, present in many commercial finite element software packages for the analysis of the viscoelastic behaviour of soils, di fiers fiom the biphasic theory in the expression of its field equations. Prendergast et al [3] have demonstrated the equivalence of three commercial poroelastic FE packages (MARC, DIANA and S WANDYNE) to FE models based on the biphasic theory. Van der Voet et al [SI and Wu er ai (61 have evaluated the solutions obtained from ABAQUS for a selection of problems, and again demonstrated their equivalence to models and analytical solutions based on the biphasic theory. To illustrate this equivalence, two examples are presented in this Appendix.
A standard experimental procedure for the determination of the material properties of a biphasic, or poroelastic tissue is the confined compression test. A cylinder of the tissue is placed in a confining well, which prevents lateral expansion, and loaded from above via a porous piston. Fluid flows out of the tissue over time and the tissue consolidates. During consolidation, the applied load is partitioned between the solid collagenous matrix and pressurisation of the interstitial fluid. At equilibrium, the applied load is camed by the solid collagenous matrix of the tissue. The relevant material properties are the aggregate compressive modulus (HA), the tissue permeability
(k) and the Poisson's ratio (v) of the collagenous matrix at equiIibrium.
S J . Fermouson Queen's University Biomechanics of the AcetabuIar Labrum
An avisymmetric mode1 of a cylindrical cartilage specimen (height = 2Smm, HA
= OSMPa, k = 7.5~10-15m4/N-s, v = 0.167) was generated, using the same finite element grid as that of Spilker et a[ [4] (ten elements equally spaced in the bottom 80 percent of the depth, five elements equally spaced in the top 20 percent). A step load, equivalent to a uniform pressure of 0.047 MPa was applied to the free-draining top surface of the specimen, and the displacement of the top surface was calculated for the period O - 10000 seconds. nie displacement results from the ABAQUS solution were compared to the displacement predicted by the biphasic theory for confined compression testing [2]. The correspondence of the two solutions is shown in Figure C. 1. Even with a relatively coane mesh. the results show excellent agreement. For the case of confined compression, the biphasic theory provides an excellent fit to experimental data.
bme (sec)
Figure C. 1 : Cornparison of the predicted consolidation displacement of a cylindrical plug of articular cartilage (h = Xmm, H,, = OSMPa, k = 7.5~1 o-'~ m4m-S. v = 0.167, applied pressure = 0.047MPa). The ABAQUS model solution matches the results of the analyticai solution based on biphasic theory, which has been previoudy shown to match experimental results [2].
S.J. Fer,won Queen's University Biomechanics of the Acetabular Labrum
------
For the second example, a confined compression stress relaxation model was -generated. In this test, a cyIindrical cartilage specimen is mounted in a confining well and the top surt'ace of the specimen is subjected to a ramp displacement through a porous piston. Again, an axisymmetric model of a cylindrica1 cartilage specimen (height =
2.5mm, radius = 3.175 mm, HA= OSMPa, k = 7.5~10- 15 m4/N-s,v = 0.167) \as generated, using the sarne finite element grid as that of Spilker el al [4]. A total Y displacement of 0.125 mm was prescribed on the top surface of the specimen over a period of 500 seconds, and then the displacement was held constant for a fùrther 500 seconds. The reaction force on the lower surface of the specimen is shown in Figure C.2, with the results of the solution fiom Spilker et al [4] shown for cornparison. Again, the two solutions show good agreement, with a 2 percent higher reaction force predicted by
ABAQUS at t = 500 seconds)
Figure C.2: Reaction force as a function of time for a confined compression stress relaxation test of a cylindncal plug of a~ticularcartilage (h = 2.5mm. r = 3.175rnm. H, = OSMPa k = 7.5~10''' m4/N-s, v = 0.167, imposed displacement = 0.12Srnm). nie ABAQUS model solution matches (within 2%) the results of the mode1 of Spilker et al [4] based on the biphasic theory.
S.J. Fermouson Queen's University Biomechanics of the Acetabular Labrum
These two fairIy simple examples demonstrate the equivalence of the biphasic
theory and the poroelastic formulations found in the commercial FE sokare Eom
ABAQ US for pro blems involvinç hydrated sofi tissues.
References
Mow VC, Kuei SC, Lai WM, Armstrong CG: Biphasic creep and stress relaxarion of anicular cartilage in compression: theory and expenkents. Jotlrnal of Biomechanical Engineering 102:73-84, 1 980
Mow VC, Zhu W, RatcliEe A: Structure and fiinction ofaticular cartilage and meniscus. In: Basic Orthopaedic Biomechanis, 1, pp 143- 198. Ed by VC Mow and WC Hayes. New York Raven Press, 199 1
Prendergast PJ, van Driel WD, Kuiper JH: A comparison of finite element codes for the solution of biphasic poroelastic problems. Proceeciings of the institution of Mechmical EngÏneers. Jorrrnal of Engineering in ,Medicine [Hj' 2 10: 13 1- 136, 1996
Spilker RL, Suh JK, Mow VC: A finite element formulation ofthe nonlinear biphasic mode1 for articular cartilage and hydrated sofi tissues including straindependent pemieability. In: ComplrfationalMethods in Bioeizgincering., 8 1-93. Ed by RL Spilker and BR Simon. New York. American Society of Mechanical Engineers, 1988
van der Voet Al?,Shrive NG, Schachar NS: Numerical modelling of articular cartilage in synoviaI joints - poroelasticity and boundq conditions. In: Cornputer Merhods M Biomechnnics and Biomedical Engineering, 200-209. Ed by J Middleton, GN Pande, and KR Williams. Swansea. IBJ Publishers, 1993
Wu JZ, Herzog W, Epstein M: Evaluation of the finite eIement sofnvare ABAQUS for biomechanical modelting of biphasic tissues, Journal of Biornechanics 3 1: 165-169, 1998
S.J . Ferguson Queen's University Biomechanics of the AcetabuIar Labrum
Appendix D: Additional Poroelastic
Finite Element Model Details
In the finite element mode1 presented in Chapter Five, a combination of poroelastic elements and interface elements were used to sirnulate the case of a relatively thick fluid layer trapped between nvo cartilage layers. The fluid layer must have the ability to support loads through the development of a hydrostatic pressure. Additionally. the fluid layer must be able to redistribute itçelf and exchange fluid with the neighbouring cartilage layers. Finally, if the fluid layer is depleted over time, the developing contact between the cartilage layers themselves rnust be modeled. ABAQUS [1,2] finite element software was used. To simulate the fluid layer, 8-node axisyrnmetric poroelastic elements
(CPE8P) we used, with a low solid matrix moduhs (0.0 1 MPa), a relatively high permeability (7.358E-2 mm/s) and a high water content (95 %). These eIements were ptaced between, and shared boundary nodes with, the poroelastic elements of the cartilage layers. The cartilage layers were modelled with 8-node axisyrnmetric poroeIastic elements (CPE8P) with a solid matrk modulus ofO. 167 MPa .a permeability of 7.358E-8 mm/s and a water content of 80%- This construction is illustrated schematically in Figure D. 1. The "fluid" etements thus created had the desired behaviour; in simple benchmark tests, loads applied to such an eIement were carried almost entirely by fluid pressures, with negligible solid stresses within the element.
S J, Ferguson Queen's University Biornechanics of the Acetabular Labrum
Figure D. 1: Elernent combination used to simulate the fluid-filied space between the cartilage surfaces. PoroeIastic ekments with Iow modulus and high water content (grey) are used to simulate water- These fom a continuous body with the poroefastic elements of the cartilage layers (white). Lnterposed between the individual nodes of the cartilage elements are node-node interface elements. The "fluid" elements can sustain a hydrostatic fluid pressure (Pr), but no solid stress, whereas in the cartilage elements, the effective stress is a combination of the fluid pressure within the tissue (Pm) and the solid stresses acting through the collagen matnx (O,,), here represented schematicalIy by springs-
With a Poisson's ratio of zero for the drained solid matrix, these fluid elements could be completely drained of their entrained water, and yet with a hi& water content, the fluid elements were essentially incompressible irnless fluid was expressed frorn the element.
Such an elernent formed a good approximation of a fluid. Due to the shared nodes between the cartilage and fluid eiements, fluid could be fkeely exchanged betwern the cartilage Iayers and the fluid layer.
Between the elernents of the cartilage layers, node-node interface elements
(WTER3A) were inserted. In the case where the fluid layer was not depleted, there was no contribution to load transfer across the joint interface by these interface elernents.
Once the fluid layer was depleted, these interface elements prevented penetration of one cartilage surface into the other (Figure D.2), and Ioads were then transferred across the interface between the joint surfaces via the local contact stresses at each interface element. The interface elernents used allowed a small amount of sliding between the two contact surfaces. Due to the relatively con-ment joint surfaces and the elasticity of the
S.J. Ferguson Queen's University Biomechanics of the Acetabular Labrum
cartilage layers, matching nodes remained geometricalIy coincident within the limits of
the interface elernents throughout the model solution.
Figure DL:Following depIetion of the fluid iayer, the node-node interface elements prevent the cartilage surfaces Eom penetrating each O ther. Loads are tram ferred via local contact stresses (G,,,~)directly across the contact zone by the interface elements.
This approximation of an entrapped fluid layer by adaptation and combination of
standard element types produced a model which e'diibited the expected reponse. While
the fluid Iayer was present, it contributed to load transfer between the opposing cartilage
Iayers through a hydrostatic pressure in the fluid. Once depleted, the fluid elements did
not contribute to load transfer; instead, load was transferred across the interface between the contacting joint surfaces by the node-node interface elements.
References
[ 1 ] AB AQUS :A BA Q US/Srancfard User's ~Manzial(version L7), Pawtucket, US A, Hi bb itt, Karisson and Sorenson, [nc,, 1 997a
[2] ABAQUS: A BAQCI'S/Strnndard Theoy iManzral (version 5.7). Pawtucket, USA, Hibbitt, Karlsson and Sorenson, Inc-, 1997b
S-J. Ferawon Queen's University Biornechanics of the Acetabular Labrurn
Appendix E: Discussion of a Proposed Method
for the Non-invasive Measurement of
Hip Joint Cartilage Layer Consolidation
Introduction
In previous experimental measurements of cartilage consolidation in diarthrodial joints, a variety OFmethods have been used to determine the time dependent changes of cartilage layer thickness in the loaded joint. Magnetic resonance irnaging (MN) has provided a non-invasive measurement in intact joints [4,5], but the resolution of MRI is curreritly limited to approximately 0.3mm, the time required for each scan limits the frequency of sampling, and thickness measurement errors are introduced with the selection of the scanning plane. A fernoml prosthesis instrumented with an ultrasound transducer has been used to provide extremely accurate measurements of cartilage consolidation [6-8, IO], but the measurements were, therefore, not of a natural joint with two opposing cartilage Iayers. A roentgenographic technique bas produced consolidation measurements in the natural joint [Il, but this technique required the insertion of marker pins and a contrast medium into the cartilage layers, and this technique is also sensitive to the selection of the imaging plane. A mechanical measurement of cartilage layer consolidation has been made using a pin following the displacement of the femoral subchondral bone sudace relative to the subchondral bone of the acetabulum 131.
However, cartilage was removed fi-om the acetabulum to provide a clearance hole for the
S.J. Ferkguson Queen's University Biomechanics of the AcetabuIar Labmm pin, thereby altering the seometry of the cartilage layer in the loaded region. A possibility for accurate rneasurement of the intact joint would be the introduction of an ultrasound transducer at the interface between the subchondral bone and cartilage of the acetabulum, but this would require a high degree of precision in placement.
A method for measuring the cartilage layer consolidation in the intact hip joint using a non-contacting three-dimensional video motion analysis system is presented. The determination of rigid-body movement from three-dimensional spatial coordinates of markers is common practise in the analysis of human motion (see, for example,
Soderkvist and Wedin [9]); however, in most cases it is oniy the relative angulation of body se-gments that is of interest in motion analysis. The proposed technique is described, sample measurements are presented, and the limitations of the method are discussed.
Method
Prior to the consolidation experiments, clusters of three reflective markers were mounted rïgidly to each joint component. One 2.5 mm steel Kirschner wire was driven into the centre of the femoral liead, and one Kirschner wire was driven into the pubis or the ischium, close to the backside of the acetabular fossa- Rigid aluminum marker holders were attached to the Kirschner wires and secured with set-screws. Using a non- contact, optical-based motion analysis system (Qualisys MacRefiex, Qualisys AB,
Goteborg, Sweden), paired video images of the reflective markers were captured on charge-coupled device (CCD) cameras, and fiom these paired images the three-
S .J. Ferguson Queen's University Biomechanics of the Acetabular Labnim
dimensionai spatial coordinates of the individual markers were calculated within the
Qualisys system software. The algorithm for calculating these spatial coordinates
required the calibration of a fixed measurement volume around the test specimen, which
aIso established a global coordinate system. Using a six-marker trame, a measurement
volume of approx imately 500 x 500 x 500 mm was established, with a global coordinate
system whose axes were aIigned with the anatomical axes of the specimen (medial-
lateral, posterior-anterior, inferior-superior).
The three-dimensional coordinates of the markers on each holder were used to
define a local coordinate system on each of the fernoral head and the acetabulum. For the
femur, a unit vector il was defined in the direction 1-2 and a unit vector i2 was defrned in
the direction 1-3. A local coordinate system was then detïned with its origin at point 1,
and unit axes:
Similarly, a local coordinate system was defined for the acetabulum, using markers 4-5-6.
The location and orientation of these local coordinate systems was arbitrary. Figure E. I
shows a schematic of the marker attachment,
Overall cartilage consolidation was defined as the change in cartilage layer thickness over time. To determine this change in cartilage layer thickness using a non-
invasive measurement, some assumptions were required about the behaviour of the overall joint when loaded. It was assumed that changes in cartilage layer thickness could
S J. Ferguson Queen's University Biomechanics of the Acetabular Labnim
Figure E. l : Marker clusters mounted on rigid steel Kirschner wires were used to define local coordinate systems (CS) on the femur (markers 1-2-3) and on the acetabuf um (markers 4-5-6).The position vector pl in the femoral CS defines the centre of the femoral head, The position vector p2 in the acetabular CS defines the centre of the acetabular socket. At the beginning of the consolidation experiment the two centres were CO-incident. Through the transformation matrices From the two local coordinate systems to the gIobaI coordinate system (xyz), the position of the two centre points were transformed into the global coordinate system.
Figure E.2: During the experirnent, the femoral head moved relative to the acetabular socket due to cartilage layer consolidation- The femoral CS (defined by 1-2-3) moved with the femur, and the position vector pl folIowed this motion. The centre of the femoral head, as located by position vector pl, displaced reIative to the centre of the acetabular socket, as located by position vector p2. Again, through the individual transformation matrices fiom local to global CS, the position of the two centres could be transformed into the global CS and the relative consolidation displacement, independent ofjoint segment rotation, could be calculated. See text for relevant assurnptions.
S.J. Ferguson Queen's University Biomechanics of the AcetabuIar Labm
be determined by measuring the relative approach of the subchondral bone surfaces of the fernord head and the acetabulum. Furthermore, it was assumed that the bone cornponents of the joint were effectively rigid, in cornparison to the cartilage layers, and that therefore the relative approach of the two joint components could be determined from the motion of the centre of the femoral head relative to the centre of the acetabular socket. If the location of the centre of the femoral head within the local femoral coordinate system was known, and the location of the acetabular socket within the local acetabular coordinate system was known, then it was possible to calculate the relative motion of these two centre points, and hence the overall cartilage layer consolidation.
The position vector pl in the acetabular coordinate systern defined the centre of the acetabular socket and the position vectorpz in the tèmoral coordinate system detïned the centre of the femoral head. During the experiment, as the fernord head rnoved relative to the acetabuIar socket, the femoral coordinate system (defined with rnarkers 1-2-3) moved with the femur. We assumed that the femur, Kirschner wire and marker holders formed a rigid body. The position vector p-, followed the motion of the femur. The centre of the femoral head, as located bypl, moved relative to the centre ofthe acetabular socket, as
Iocated by the position vectorpl. In order to calculate the relative displacement of the femoral head within the acetabular socket, to determine the overall cartilage Iayer consolidation, it was necessary to express both centre points in a common coordinate system, the globai coordinate system (Figure E.2). The relative displacement was then simply the vector di fference of the position of the two centre points in the global coordinate system, which was aligned with the anatomical axes of the body.
S.J. Ferguson Queen's University Biomechanics of the Acetabular Labrum
Transformation of the position vectors from local to global coordinate systems
was straightforw*ard. For two arbitrary coordinate systems, X-Y-Z and .y--r,the
relationship between the X-Y-Z coordinates and the x-y-z coordinates of any point P is
civen by: C
where Tt: TI.and TZ are the X: Y and Z cornponents of the translation vector at the origin
of the -r-v-z system and [RI is the rotation matrix:
and 1, J and K are the unit axes of the coordinate system X-Y-Z and i, j and k are the unit
ayes of the -Y-y-z coordinate system.
Prior to the consolidation experiments, it was necessary to determine the location
of the centre of the kmoral head and of the acetabular socket in their respective
coordinate systems. The human hip joint is a remarkably spherical, alrnost congruent joint, with deviations from sphericity of the cartilage layers of less than 100pm, and
congruency of the hvo joint surfaces to the micron level, according to the study of Tepic
[IO]. Therefore, it \vas reasonable to assume that any rotation of the unloaded joint could
be described as a rotation about a single point, the joint centre. To determine the joint
centre, motion data was gathered while the femur was moved through a motion of
S.J. Ferguson Queen's University Biomechanics of the AcetabuIar Labrurn
approximately * 30' of flexion/extension and * 30" of adduction/abduction. The spatial coordinates of the six markers were transformed into the acetabular coordinate systern and the coordinates of the femoral markets were fit individually to the equation of a sphere, using a least-squares criteria to find the best fit for the sphere radius and centre point in the acetabular coordinate system. This determined the position vector pl for the centre of the acetabular socket in the acetabular coordinate system. To find the position vectorp? for the centre ofthe femoral head in the fernoral coordinate system, the position vector p, was transformed to the femoral coordinate system for each time step of the motion data gathered while rotating the joint, and the average of al1 these calculated vectors kvas taken.
Three sets of tests were made to evaluate this proposed method, First, the measurements were simulated using computer generated data to evaluate the influence of marker holder size, number of markers, Kirschner wire len-& (position vector Lena&), and the addition of noise to the spatial coordinates calculated by the Qualisys system.
Second, the consolidation experiment was simulated using a metaVpolyethylene hip prosthesis mounted on a translating stage with a micron-resolution linear actuator (mode1
VP30-25, NewportMicro-Controle S.A., Evry, France). Finally, measurements of cartilage layer consolidation were attempted duhg the in vim whole joint experiments described in Chapter Seven, using this method. The computer routines used to calculate consolidation displacement are included at the end of this Appendix.
S.J. Ferguson Queen's University Biomechanics of the Acetabular Labntm
Computer Simulation
Computer generated data for the initial flexiodextension - adduction/abduction movements and the subsequent joint consolidation was used to evaluate the accuracy of the proposed method for determining cartilage Iayer thickness changes. An ideal data set was created, with values deterrnined to the nearest micron, as the data from the Qualisys system is given to the nearest micron. The two joint centres initially coincided at the slobal onsin. The imposed motion of the femoral component was a 300 pm linear ramp displacement along the global Z-ais, together with a rotation of 30° about the global 2- ais, to test the influence of rotation of the virtual markers on the calculated relative displacement of the joint centres. A variety of different test configurations were simulated, with norrnally distributed noise added to the spatial coordinates of the markers, and with di fferent marker holder sizes, different marker arrangements and different femoral position vector Iengths. The test configurations are summarised in
Table E. 1:
Spatial averaging of the coordinates of five markers arranged in a cross pattern reduced the noise level of the three ''virtual" markers used in subsequent calcuIations by approximately haIf. The influence of the different testing configurations on the accuracy of the measurements is shown in Figures E.3 - E.6. The quality of the spherical fit for the flexiodextension - adductiodabduction motion is shown in Figure EL This spherical fit of the motion data was used to determine the two position vectors pl and pi.
Noise imposed on the data increased the error in the prediction of these two position
S J. Ferguson Queen's University Biomechanics of the Acetabular Labrurn
vectors, as did an increase in the length of the position vectors (effectively, the length of the 'tirtual" Kirschner wire), or a decrease in the size of the rnarker holders (Figure E.4).
The addition of a relatively small amount of noise to the system resuIted in a relatively large error in the calculation of relative joint displacement (Figure E.5, E.6), but these errors could be decreased by pre-filtering the raw data of marker spatial coordinates before processing.
Table E.1: Test configurations simuIated
Number of Marker Noise Added Position Test hl arkers Spacing * to Data Vector (Max, Std. Dev.) Lena&
20mm - L 40pm. 23pm 1 OOmm
4Ornm - X 40pm, 23pn 150nim
40mm - L 40pm, 23pm 100rnrn
3Omm - L 40pm. 23pm 1 OOmm
40mm - X 40pm, 23pm l OOmm
40mm - X 30pm, 17pm 1 OOmm
40mm - X Opm. Opm l OOmm
*L Three markers amnged in an L, as indicted in Figure C.1 X Five markers arranged in a cross. Three "virtual" markers were created for analysis by spatial averaging of the coordinates of the five actual markers.
S .J Ferguson Queen's University Biomechanics of the Acetabular Labrum
Qualrty of Spherical Fit
test mnfguration
Figure E.3: Standard deviation of the sphere radii calculated for al1 data points during tlexion/extension - adduction/abduction motion for the calculation of joint centres.
Errors in Position Vectors
MO
test confgufation
Figure E.4: Absolute errors of the caiculated position vectors for the joint centres in the acetabular coordinate system @,) and in the femoral coordinate system (p2). Error bars indicate the variation in the centres calculated fiom each of the markers for the position vector pl, and the variation in centres calculated with the fernord position for the position vector p2.
S.J. Ferguson Queen's University Biomechanics of the AcetabuIar Labm
- -
Maximum Emrs in Consolidation Calculation
test configuration
Figure E.5: Maximum errors in calculated consolidation displacement tor computer generated data
Average Enor in Consolidation CaIculation
.'ZZJ 0 unfiltered 1
CDEFG test configuration
Fi-me E.6: Average errors in calculated consolidation displacement for computer generated data. Error bars indicate one standard deviation.
SJ. Ferguson Queen's University Biomechanics of the Acetabufar Labnun
Evaluation of System Accuracy
The displacement of single markers mounted on a stepper-motor driven linear stage with micron-level accuracy was measured with the Qualisys system to provide an
indication of the expected spatial accuracy of such measurements. In a similar measurernent volume as that used for the final consolidation testing, and with cameras placed 2m fiom the stage and 2m apart, step displacements of 200pm and 400ym were irnposed and measured- The measured displacement (sampled at 6OHz) fiom one tnaI is shown in Figure E.7, before and after filtering with a simple 30-point rnoving average.
Errors of Iess than ZOp, after filtering, were obsewed in al1 trials.
Marker Displacement
tirne (sec)
Figure E.7: Displacement of an individual reflective marker, as measured by the Qualisys system, Markers were mounted on a micro- controlled linear stage with stepper motor, and disptacements of 200 pm and 400 prn were imposed.
S J. Ferabuson Queen's University Biomechanics of the Acetabular Labrum
A second test of system accuracy was the simulated consolidation experiment using a hip prosthesis as a phantom for the real joint. Acetabular markers were mounted to the fixed base of the linear stage and femoral markers were mounted to the prosthesis, which was attached to the rnoving component of the stage. With no a priori knowledge of the correct location of the position vectors pl and pz in the prosthesis, it waç not possible to evaluate the absolute error in the calculation of these vectors, as was done with the computer generated data set. However, it was possible to calculate the quality of the spherical fit of position data during the determination of the joint centres, and also the variation of the calculated centres fiom each of the three femoral markers used (shown as error bars in Figure E.4 for the computer generated data). The standard deviation of the sphere radii calculated was, on average, 65 Pm- The variation in position vectors calculated from each of the three markers was 56 pm for vectorp! on the acetabulum, and 165 pm for the vector pz on the femur over the range of femur motion.
A series of increasing displacements were imposed using t. manual positioning controls of the linear stage. The calculated relative displacement of the two joint centres is compared in Figure E.8 to the imposed displacement.
S.J. Ferguson Queen's University Biornechanics of the Acetabular Labrum
Consolidation Simulation
1 -calculated displgcernent -irnposed displacernent I I
-1 5 1 I O 100 200 3W 600 500 MX] time (sec)
--
Figure E.8: ReIative displacement of the femoral head with respect to the acet bular socket, with displacements imposed via a micro-controlled linear stage with stepper motor. The solid horizontal Iines indicate the imposed displacement. For small displacements (less than 700 p), the agreement is good benveen the displacement calculated by the proposed method and the actual displacement. For Iarger displacements, the error is greater than 100pm.
Hip Joint Consolidation Measurements
The proposed method was used to rneasure cartilage layer consolidation during in vitro hip joint experiments. Again, it was not possible to evaluate the absolute error in the calculation of the position vectors pl and pz ,as was done with the computer generated data set. However, it was possible to calculate the quality of the spherical fit of position data during the determination of the joint centres, and also the variation of the calculated centres hmeach of the three femoral markers used. The standard deviation of the sphere radii calculated was, on average, 273 Pm. The variation in position vectors
S.J. Ferawon Queen's University Biomechanics of the Acetabular Labrum
-- calculated from each of the three markers was 205 prn for vectorpr on the acetabulum, and 1440 p for the vectorpr on the femur over the range of femur motion.
As an example, the raw Z-displacement (vertical) of two markers, as measured by the Qualisys system, during consolidation testing is show in Figure E.9. The shape of the displacement curves is similar to that of the displacement curves recorded by the linear variable diFferential transformer (LVDT) of the MTS testing kame.
Raw Qualisys Marker Data
-acetabulum O0 - femur
E -E
O lm Zoo0 30m 4000 time (sec)
Figure E.9: Individual marker displacement (vertical direction) as determined by the Qualisys video motion analysis system. Zero displacement corresponds to the marker position at the be$nning of load application. The shape and magnitude of the displacement curves are consistent with the overall joint compression rneasured by the LVDT' of the MTS testing kame.
When this spatial coordinate data was then used to calculate the consolidation of the cartilage layers, the results were poor. The vertical consolidation of two joints under static loading is shown in Figure E. 10. For one joint, the calculations predict no
S.J. Feraouson Queen's University Biomechanics of the Acetabuiar Labrum
consolidation, on average- For the second joint, the calcufations predict a cartilage layer
consolidation greater than the overall joint compression measured by the LVDT of the
MTS test fiame.
Calculated Consolidation Displacement
1.s
-0.5 1 O 1000 2m 3000 4000 tirne (sec)
Figure E. IO: Cartilage layer consolidation for two hip tests as calculated from Qualisys marker data. Spatial coordinate data was first filtered, and then processed as described in the method section. The results of the calculations are not consistent with the measured compression of the whole joint.
Discussion
The cartilage consolidation measurements taken during loading of whole hip joints in vitro did not provide data with the accuracy predicted frorn preliminary computer and experimental simulations. Error levels were higher than expected at al1 steps of the calcuiation, although the leveI of system noise (changes in relative distance between two rigidly connected markers) was within the 20 -40 pm range expected. The
S.J- Ferguson Queen's University Biomechanics of the Acetabular Labnrm overall accuracy of the proposed method is not yet adequate for measuring relatively srnall changes in cartilage layer thickness (< 500 pm).
One possible source of error was the assumption that the bony structures of the acetabulum and femur were rigid, as were the connections through the Kirschner wires to the rnarker holders. The calculations of consolidation displacement assumed that the entire bony structure of each joint element, together with the video markers, moved as a rigîd body- Any relative motion between the subchondral bone surface and the marker hotder would lead to an erroneous calculation of cartilage layer consolidation. To minimise this possibility, the Kirschner wire for the femoral markers was inserted, through a clearance hole, into the centre of the femoral head. With Iess bone stock avaiIab1e in the pelvis, and the necessity of Iocating the marker holder above the surface of the fluid bath, it was not possible to optimally place the Kirschner wire For the acetabular markers directly under the subchondral bone of the acetabular socket.
However, Dalstra et ni (21 have shown, with an experimentally validated tïnite element mode1 of the entire pelvis, that the stresses in the cortical shell of the acetabulum, for an applied Ioad of approximately 0.75 - I times bodyweight, are several orders of magnitude less (0.04%) than the rnodulus of the cortical bone itself. One can infer fiom this that the strain levets within the acetabulum and femur, and therefore the deformation of the bony structures, are negligible, for load levels of the same magnitude as those used in the current esperiment,
A more likely source of error in the results obtained fiom the in vitro consolidation experiments was the inaccuracy of the Qualisys systern itself Following the disappointing and somewhat puzzling, results of the experiment, further evaluations
S.J. Ferguson Queen's University Biomechanics of the Acetabular Labrum
of the Qualisys system were made- it was found that: (1) the system is extremely
sensitive to the angular orientation of the marker holders with respect to the viewing
plane of the cameras (2) the static calibration fiames and the factory calibration technique
of Qualisys were flawed and (3) the method used by Qualisys for determining and
reporting system accuracy \vas not representative of the actual system errors. Tests conducted after the in vitro experiments, using a series of rotating rigïd bodies, revealed errors of up to 400ym depending on the orientation ofthe markers. The previous evaluation of the system accuracy, using markers mounted to a precision linear actuator, did not reveal this sensitivity as the markers were mounted in the plane of the cameras' view. Tests of systern accuracy based on calculations of the relative displacement of two rigidly connected rnarkers, as defnned by Qualisys, reflected more the resolution of the system than the actuaI spatial accuracy of individual marker positions. While the relative distance between hvo rigïdly connected rnarkers never varied by more than *40pm, it is probable that the determination orthe spatial location of those ttvo markers, and the actual distance calculated, had muich larser errors. Further discussions with the engineering staffat Qualisys have resulted in the retum of the calibration cames to
Qualisys. The poor quality of the rnarkers used on the calibration frame, and the caiibration technique employed by Qualisys resulted in a non-linear distortion of the measurement volume. Data obtaimed from this poorly scaled measurement volume could not be re-scaled afier testing, as pnecise spatial reference measurements would have been required at the time of calibration on the MTS testing fiame, with the calibration fkame in place. With no knowledge of these inherent errors of the systern, this was not done. The current status of the Qualisys system is that the Company no longer proposes the
SJ. Ferguson Queen's University Biomechanics of the Acetabutar Labnun calibration frame and technique that they had originally recommended, and are working towards developing a more precise, dynamic calibration method. That the originaI measurements made to evaluate the systern accuracy produced excellent results seems now to be more a result of good fortune than system quality. Nevertheless, these errors could theoretically be eliminated through better design of the calibration procedures and retèrence fiames. A static calibration fi-ame is not inherently inacurate, and the static calibration frame could be improved through the addition of redundant markers, with appropriate changes to the calibration aigorithm to account for these extra markers. The resolution of the current CCD cameras of the Qualisys system is more than adequate for the hi& lever of accuracy that this proposed consolidation measurement method would require. Further substantial improvements in system accuracy could me made by incorporating additional, redundant cameras, While this alone may be suKcient to reach the Ievel of accuracy required for the proposed measurement technique, there is a significant cost associated with each additional camera and related video processing equipment.
An alternative method for accurately measuring the relative rotation and displacernent of two joint components is that of Wang et al [Il]. Their apparatus consisted of an orthogonal frarne containing a soiid cube, the motion of which was followed by sis spring-loaded LVDTs. This apparatus provided measurements with a translation accuracy of 0.0 I mm and a rotation of 0-04"; the addition of a seventh, redundant LVDT improved the accuracy of some measurements by as much as 50%.
However, this apparatus was designed for measuring the motion of the sacroiliac joint, with rotations of less than 2" on average, although the apparatus was caiibrated for
SJ. Fer,won Queen's University Biomechanics of the Acetabular Labrum rotations of up to 8O [1 Il. In our tests, the range of rotation necessary to accuntely determine the centre of the hip joint was approximately 20 - 30" of flesiordextension or adduction/abduction. It is possible that modifications to the design of Wang et al would allow this technique to be applied for the measurernents described here. The weight of their relatively compact apparatus was approximately 1 kg. The design of the apparatus required a ngid connection from the two components of the device to the joint se-ments being measured. One advantage ofthe video based motion analysis system was the light weight of the markers, allowing the use of relatively thin Kirschner wires for connection to the femur and to the thin cortical bone of the acetabulum, which provides îàirly limited opportunities for rigid placement of such hardware. Also, the location and orientation of the hvo video markers holders were not constrained, which was an important consideration when working within an experimental setting where the joint components were submerged in a bath.
While the proposed rnethod could not be used in the current in vitro consolidation experiments, it would seem that the method could provide accurate measurements of joint consolidation, or cartilage layer compression over tirne. With carefùl pIacement of the
Kirschner wires, even usins a fluoroscope to place the tips within the subchondral bone layer, and with a proper calibration of the measurement volume and spatial coordinate data from the video motion analysis system, the accuracy levels predicted in the computer simulations should be attainable.
S.J. Ferguson Queen's University Biomechanics of the Acetabular Labrum
MATLAB and EXCEL Procedure
for Calculating Joint Consolidation
frorn 3D Marker Data
MATLABroutine to transform spherical motion marker data into pelvic coordinate system
% Endjoint-center.m Y0 % load in MacReflex data Eorn initial measurements to determine % joint centre relative to femur and acetabulum marker triads % load data-txt for m= 1:size(data, 1 ) Y0 % assig data to individual markers % markers is an t x n*3 matrix where n is the number of % markers measured by the MacReflex system in one time hme Y0 for n= 1 :6 markers(n, 1 :3)=data(m,(n*3-2):n*3); end % al1 data is in MacReflex global coordinates % transforrn it into the pelvis coordinate system % marker matrix is 6x3 for export to transform function % first three rows are marker triad on CS to be transtormed into % second three rows are marker triad on CS to be transformed from marker = [markers(4:6,I:3);0,O10;1,0,0;0,1 ,O]; % calculate transforrn matrix T = transform(rnarker): % transforrn MacReflex coordinates fiom global into pelvis coordinate fi-ame for n= 1 :6 p=[ 1,markers(n, 1 :3 )]; p'p'; p=T*p: newcoords(m,(n*3-2):n*3)=p(2:4)'; end end Save .\globqelv.txt newcoords -ascii
S.J. Ferguson Queen's University Biomechanics of the Acetabutar Labrum
Read data into an Excel spreadsheet and use the "Solver" fùnction to find the Ieast-squares best- fit sphere for the data, CalcuIate the joint centre in the pelvis coordinate system:
Equation of a sphere Fitting MacReflex data to find center of the joint in pelvis coordinate syste 1 Use "Solver" to find best-fit centre and radius
L I 103.1 46 average r 157.98904 j -82.024 stdev r 0.3661 398 surn RA2 138.28217
i, j, k, r} varied by Solver to minimize (sum RA2) I{- - Paste MacReflex Data here
The following MATLAB routine calculates the joint centre in the Femoral coordinate system, and then ca1crilates the total joint consolidation (relative motion of the two joint centres) fi-om the experimentd data:
% conso1.m Y0 % now convert the joint center tiom pelvis coordinates into femur coordinate % later this is averaged to provide an average position vector for the %joint center in femur coordinates Y0 % load in al1 MacRefles coordinates. now in Pelvis coordinâtes Y0 load globqe1v.txt pelvis- cente~input('Joint Center in Pelvis CS? '); center-sum=[O O O]; for m= 1 :size(globqelv, 1 ) % assign data to individual markers % markers is an 1 x n*3 matrix where n is the number of % markers measured by the MacRetlex system in one time fiame for n= 1 :6 markers(n.l:3)=gIobqelv(m,(n*3-2):n*3); end % al1 data is in pelvis coordinates % transform it into the femur coordinate system % markeï matrix is 6x3 for export to transform fùnction % first three rows are marker triad on CS to be transfonned into (femur) % second three rows are marker triad on CS to be transfomed from (pelvis) marker = [markers( 13.1 :3);markers(4:6,1:3)];
SJ. Fer,ouson Queen's University Biomechanics of the Acetabular Labrum
% calculate transform matrix T = transfonn(marker); % transform joint center coordinates into tèmur coordinate frame
end
Save ,ken-fem-txt femur-center all-femur-center -ascii Y0 % now read in MacReflex data fiom the consolidation test % use the first time step to estabtish the three-virtual % markers defining thé joint center coordinate frarne in both % fernur and pelvis cooidinate systems % load consoldata.txt for n=l:6 markers(n, 1 :3 )=consol-data( 1 .(n*3-2):n*3); end % % now for the tün part % for each time stc$ of the consolidation test % calculate the dis&tcementbetween the % pelvis center and Femur center Y0 for m= 1 :size(consol-data, 1 ) for n=I :6 markers(n, 1 :3)=consol~data(m.(n*3-2):n*3); end % al1 data is in MacReflex global coordinates % transfomi it into the pelvis coordinate systern % rnarker matrix is 6x3for export to transform function % first three rows are rnarker triad on CS to be transformed into ?4 second three rows are marker triad on CS to be transformed From marker = [markers(4:6,1:3);0.0,0; 1,0,0;0.1 ,O]; % calculate transform rnatrix T = transform(marker); % transform MacReflex pelvis center from pelvis CS to global CS p=[ I ,pelviscenter]; p-p' ; p=inv(T) *p ; pelvis-center_global=p(2:4); % now do the same for the tèrnur center marker = [markers( 1 :3,1:3);0,0,0; 1,0,0;0,1 ,O]; % calculate transform matrix T = transform(rnarker); Oi'o transform MacReflex femur center fiom femur CS to global CS
S.J. Ferguson Queen's University Biomechanics of the Acetabular Labrum
p=inv(T) *p; femur-center-g Io bal=p(2:4); % wete the results to a matrix newcoords2(rn, 1 :3)+emur-center-global'; newcoords2(m,4:6)=peIvis~center~global'; displacement(m, 1 :3)=(femur- centerglobal-pelvis-center~lobal)'; end Y0 % these displacements are in the MacReflex coordinate system % not in th; anatomical directions, as the pelvis is tipped % in the loading hme for somo tests % time to do the final conversion Y0 Y0 % positive X displacement is a lateral displacement % positive Y displacement is anterior for right hip, posterior for left hip! % positive Z displacement is a superior displacement Y0 p-tilt = input('Pe1vic Tilt (degrees)? '1; hip = input('Leti( 1 ) or Right (2) hip tested? '); ptilt = ptilt*pil l8O; if hip = 1 rnarker = [0,0,0; 1.0,0;0,1 ,O;O,O,O;-cos(p-tilt),O.sin(ptilt);O, 1,O]; else hip = 2 marker = [0,0,0: 1 .O,O;O1 1 ,O;O,O.O:cos(p~tilt),O,sin(p~tilt);O,- 1 .O]; end
for m = 1:size(displacement, 1 ) p = [ 1.displacement(rn, 1 :3)]: p = p'; p = T*p; displacornent(m. 1 :3 ) = ~(24)'; end Save .\consoI-disp-txt displacement -ascii
?'O ?'O transformm % function T = transform(marker) 0/o calculate transformation matrix between two bodies % transformation matrix transforms points in the CS % associated with B into the CS associated with body A % marker is a 6x3 rnatrix of rnarker triad coordinates on % A (rows 1-3) and B (rows 4-6) Y0 % assign data to individual markers % 1,J.K on A t=[marker(2, I :3)-marker(l,l:3)]; t=I/norrn(L); J=[marker(3,l:3)-marker( 1. I :3)]; J=J/norm(J) ; K=cross(l.J):
S.J. Ferguson Queen's University Biomechanics of the Acetabular Labrum
J=cross(K,I ); % ij,kon 8 i=[marker(5,1:3)-marker(4, t :3)]; i=i/norm(i); j=[marker(6,1:3)-marker(4,1:3)]; j=j/norm(j); k-oss(i 2); j=cross(k, i); % rotation matrix to convert arbitrary position vector fÏom B to A R=[surn( i.*i) sum(j.* l) sum(k,* 1) sum(i-*J ) surn(j-*J) sum(k*J) sum(i.*K) sum(j.*K) sum(k.*K)]; % translation vector fiom B to A % this has to be expressed in the A coordinate system t=[marker(4,1:3)-marker( l,l:3)]; % this is in B coordinates L-L' ; t=R*t; % output is 4x4 transform matrix T T=[I 0 0 O;t,R];
S.J. Ferguson Queen's University Biomechanics of the Acetabulax Labrum
References
Armstrong CG, Bahrani AS, Gardner MA: in-Vitro measurement of articular cartilage deformations in the intact human hip joint under load, Jorrrnal of Bone undJoinr Szrrgery 6 1-A:744-755, 1979
Dalstra M, Huiskes R, van Erning L: Development and validation of a threedimensional finite element mode1 of the peivic bone. Jorrrnal of Biomecltanicul Engineering 1 17273- 278. 1995
Day WH, Swanson SA, Freeman MA: Contact pressures in the loaded human cadaver hip. Jozrrnal of Bone and Joint Srrrgen; 57-A:302-3 1 3, 1 975
Eckstein F, Tieschky Ml Faber SC, Haubner M, Kolem H, Englrneier K-H. Reiser M: Effect of physical exercise on cartilage volume and thickness in vivo: MR imaging study. Radiologz. 207:243-238, 1998
Herberhold C, Starnrnberger T, Faber S, Putz R Englmeier K-H, Reiser M, Eckstein F: An Mr-based technique for quanti@ing the deformation of articular cartilage during mechanical loading in the intact cadaver joint. hfagnetic Resonance in Medicine 39:843 - 850, 1998
Macirowski T, Tepic S, Mann RW: Cartilage stresses in the human hip joint- Journal of Biomechanical Engineering 1 16: 10-1 8, 1994
Rushfeldt PD. Mann RW, Harris WH: lmproved techniques for rneasuring in vifro the geometry and pressure distribution in the human acetabulum - 1, ultrasonic measurernent of acetabular surfaces, sphericity and cartilage thickness. Jotirnal of Biomechanics 14:353- 360, 1981a
Rushfeldt PD, Mann RW, Harris WH: Improved techniques for measuring Nt vitro the geometry and pressure distribution in the human acetabulum - II. instmmented endoprosthesis measurement of articular sufiace pressure distribution. Jozrrnal of Biomechanics I4:3 15-323, 198 1b
Soderkvist 1, Wedin P-A: Determinhg the rnovements of the skeleton using well- configured markers, Jolrrnai of Biomechanics 26: 1473- 1477, 1993
Tepic S: Dynamics of and enrropy prodrrction in rhe cartiIage lqers of the synovialjoint., Sc-D. Thesis, Massachusetts institute of TechnoIoa. 1982
Wang M, Bryant JT, Dumas GA: A new in vitro measurement technique for srnaII three- dimensional joint motion and its appkation to the sacroiliacjoint. A&edical Engineering and Physics l8:495-50 1, 1996
S.J. Ferguson Queen's University