Matrix-Induced Autologous Chondrocyte Implantation for Articular Cartilage Injury: Biology, Histology, and Clinical Outcomes

MATRIX-INDUCED AUTOLOGOUS CHONDROCYTE IMPLANTATION FOR ARTICULAR CARTILAGE INJURY: BIOLOGY, HISTOLOGY, AND CLINICAL OUTCOMES

Craig Robert Willers BSc(H1), M(Med)Sc

This thesis was submitted as fulfillment of the requirements for the degree of

DOCTOR OF PHILOSOPHY IN MEDICAL SCIENCE 2008

The research presented in this thesis was performed at The Centre for Orthopaedic Research, School of Surgery, The University of Western Australia, Queen Elizabeth II Medical Centre, Nedlands, Western Australia.a i TABLE OF CONTENTS

NOTE: If viewing electronically (PDF), please click heading to link to associated page

DECLARATION vi

ACKNOWLEDGEMENTS vii

DEDICATION viii

PUBLICATIONS ix

ABBREVIATIONS xi

LIST OF FIGURES xiv

LIST OF TABLES xxvi

THESIS ABSTRACT xxviii

CHAPTER 1: GENERAL INTRODUCTION & THESIS OBJECTIVES 1.1 GENERAL INTRODUCTION 1

1.2 THESIS OBJECTIVES 5

CHAPTER 2: INTRODUCTION TO CARTILAGE REPAIR TECHNIQUES Thesis publication #1: Willers C, Partsalis T, Zheng MH. Articular cartilage repair: procedures versus products. Expert Rev Med Devices. 2007; 4(3): 373-92.

2.1 STATEMENTS OF AUTHORSHIP AND PERMISSION 9

2.2 SUMMARY 11

2.3 THE BURDEN OF ARTICULAR CARTILAGE INJURY 11

2.4 NATIVE ARTICULAR CARTILAGE REPAIR 12

2.5 CARTILAGE REPAIR PROCEDURES 12

2.5.1 Arthroscopic Lavage and Debridement 13

2.5.2 Microfracture 13

2.5.3 Osteochondral Autografting (Mosaicplasty) 14

2.6 CARTILAGE REPAIR PRODUCTS 15

2.6.1 Carbon Fibre Implantation (CFI) 15

2.6.2 Periosteal Autologous Chondrocyte Implantation (ACI-P) 16

2.6.3 Collagen-Covered ACI (ACI-C) 17

2.6.4 Matrix-Induced ACI (MACI) 17

2.6.5 Hyaluronic Acid ACI 18

2.6.6 Collagen Gel ACI 18

2.6.7 Allogenic Chondrocyte Implantation 19

2.6.8 Stem Cell-Based Cartilage Repair 19

2.7 COMPARISON OF EFFICACY BETWEEN CARTILAGE 20

REPAIR PRODUCTS AND PROCEDURES

2.8 IMPACT OF CONCOMINANT JOINT SURGERY 21

2.9 EXPERT COMMENTARY 22

2.10 FIVE-YEAR VIEW 22

2.11 KEY ISSUES 24

2.12 REFERENCES 24

CHAPTER 3: MOLECULAR BIOLOGY OF THE MACI TECHNIQUE

Thesis publication #2: Kirilak Y, Pavlos NJ, Willers CR, et al. Fibrin sealant promotes migration and proliferation of human articular chondrocytes: Possible involvement of thrombin and protease-activated receptors. Int J Mol Med. 2006 Apr; 17(4): 551-8.

3.1 STATEMENTS OF AUTHORSHIP AND PERMISSION 31

3.2 ABSTRACT 38

3.3 INTRODUCTION 38

3.4 MATERIALS AND METHODS 39

3.5 RESULTS 40

3.6 DISCUSSION 43

3.7 REFERENCES 44

iii CHAPTER 4: SHEEP MACI: REPAIR ASSESSMENT BY CONFOCAL

ARTHROSCOPY

Thesis publication #3: Jones CW, Willers C, Keogh A, Smolinski D, Fick D, Yates P,

Kirk TB, Zheng MH. Matrix-induced Autologous Chondrocyte Implantation (MACI®) in

Sheep: Objective Assessments Including Confocal Arthroscopy. Journal of Orthopaedic

Research, 2007.

4.1 STATEMENTS OF AUTHORSHIP AND PERMISSION 46

4.2 ABSTRACT 51

4.3 INTRODUCTION 51

4.4 MATERIALS AND METHODS 52

4.5 RESULTS 55

4.6 DISCUSSION 60

4.7 REFERENCES 61

CHAPTER 5: CLINICAL MACI: BIOLOGY AND HISTOLOGY

Thesis publication #4: Zheng MH, Willers C, Kirilak L, Yates P, Xu J, Wood D,

Shimmin A. Matrix-induced autologous chondrocyte implantation (MACI): biological and histological assessment. Tissue Eng. 2007; 13(4): 737-46.

5.1 STATEMENTS OF AUTHORSHIP AND PERMISSION 63

5.2 ABSTRACT 67

5.3 INTRODUCTION 67

5.4 MATERIALS AND METHODS 68

5.5 RESULTS 70

5.6 DISCUSSION 72

5.7 REFERENCES 75

iv CHAPTER 6: CLINICAL MACI: REVISED AND FAILED HISTOLOGY

Thesis publication #5: Willers C, Stoffel K, Zheng MH. Histological assessment of revised and clinically failed matrix-induced autologous chondrocyte implantation.

Provisionally accepted into Osteoarthritis and Cartilage, May 2008.

6.1 STATEMENTS OF AUTHORSHIP AND PERMISSION 77

6.2 ABSTRACT 80

6.3 INTRODUCTION 81

6.4 MATERIALS AND METHODS 82

6.5 RESULTS 86

6.6 DISCUSSION 88

6.7 FIGURES 94

6.8 TABLES 99

6.9 REFERENCES 103

CHAPTER 7: CLINICAL MACI: FUNCTIONAL AND STRUCTURAL

ASSESSMENT

Thesis publication #6: Robertson WB, Willers C, Wood DJ, Linklater JM, Zheng MH,

Ackland TR. Matrix-induced autologous chondrocyte implantation (MACI) at two years: MRI and functional evaluation. Accepted to The Knee, February 2008.

7.1 STATEMENTS OF AUTHORSHIP AND PERMISSION 107

7.2 ABSTRACT 112

7.3 INTRODUCTION 113

7.4 MATERIALS AND METHODS 115

7.5 RESULTS 119

7.6 DISCUSSION 122

7.7 FIGURES 127

7.8 TABLES 133

7.9 REFERENCES 135

v CHAPTER 8: CLINICAL MACI: PATIENT SATISFACTION SURVEY

Thesis publication #7: Willers C, Zheng MH. Matrix-induced Autologous Chondrocyte

Implantation (MACI®): A Retrospective Survey of 202 Cases.

8.1 STATEMENTS OF AUTHORSHIP AND PERMISSION 140

8.2 ABSTRACT 143

8.3 INTRODUCTION 144

8.4 MATERIALS AND METHODS 146

8.5 RESULTS 148

8.6 DISCUSSION 150

8.7 TABLES 155

8.8 REFERENCES 159

CHAPTER 9: GENERAL DISCUSSION

9.1 GENERAL DISCUSSION 162

9.2 CONCLUSION 169

BIBLIOGRAPHY: 170

vi DECLARATION

This declaration functions to certify that all work documented within this PhD thesis was performed by the candidate, Craig Willers, unless indicated otherwise. This thesis is submitted for the degree of Doctor of Philosophy at the University of Western Australia, and has not been submitted for any other qualifications at other institutions.

CANDIDATE

Craig Robert Willers

CO-SUPERVISORS

Professor David Wood

Dr Nathan Pavlos

SUPERVISOR

Professor Ming-Hao Zheng

vii ACKNOWLEGDEMENTS

I would like to thank everyone and anyone who has helped me over the course of this PhD, whether inside or outside the laboratory.

In particular I would like to sincerely thank my supervisor Professor Ming-Hao Zheng, whose continuous guidance, support and enthusiasm has been essential for the successful evolution of this PhD. Thanks for the encouragement Zheng! Great appreciation must also be given to my co-supervisors Dr Nathan Pavlos and Professor David Wood.

I would not have achieved much of the data presented herein without the collaboration of my fellow researchers and publication co-authors. In particular, I would like to thank Dr Chris Jones from the School of Mechanical Engineering UWA, Brett Robertson from the School of Human Movement and Exercise Science UWA, and Lyn Kirilak from the Centre for Orthopaedic Research UWA, for their interest in my MACI research and significant contribution to my thesis.

Of course, I must salute the rest of the Orthopaedics gang who have provided academic and moral support through my studies and made the work environment a little more amusing (particularly ‘Tolerant Tamara’ who had the unfortunate pleasure of sharing an office with me).

Also, thanks to Slavica Pervan for her invaluable assistance with the processing and staining of my histological samples, and her infectious giggle.

viii DEDICATION

I would not be where I am today without the love and patience of my family. I would like to dedicate this thesis, and all it represents, to my family for always being there when needed, regardless of the colour of mind. Thank you for all the assistance you have given me – nothing is taken for granted, and nothing is forgotten. Hopefully one day I can give back a little of what you have given to me.

YÉÜ Wtw? `âÅ? tÇw a|vÉÄx ix PUBLICATIONS – CONTRIBUTING TO THIS THESIS

(1) Willers C, Stoffel K, Zheng MH. Histological assessment of revised and clinically failed matrix-induced autologous chondrocyte implantation. PROVISIONALLY ACCEPTED INTO ‘OSTEOARTHRITIS AND CARTILAGE’ AUGUST 2008.

(2) Robertson WB, Willers C, Wood DJ, Linklater JM, Zheng MH, Ackland TR. Matrix-induced autologous chondrocyte implantation at two years: MRI and functional evaluation. ACCEPTED TO ‘THE KNEE’ FEBRUARY 2008.

(3) Jones CW, Willers C, Keogh A, Smolinski D, Fick D, Yates P, Kirk TB, Zheng MH. Matrix-induced Autologous Chondrocyte Implantation (MACI®) in Sheep: Objective Assessments Including Confocal Arthroscopy. Journal of Orthopaedic Research. 2007.

(4) Willers C, Partsalis T, Zheng MH. Articular cartilage repair: procedures versus products. Expert Rev Med Devices. 2007; 4(3): 373-92.

(5) Zheng MH, Willers C, Kirilak L, Yates P, Xu J, Wood D, Shimmin A. Matrix- induced autologous chondrocyte implantation (MACI): biological and histological assessment. Tissue Eng. 2007; 13(4): 737-46.

(6) Kirilak Y, Pavlos NJ, Willers CR, Han R, Feng H, Xu J, Asokananthan N, Stewart GA, Henry P, Wood D, Zheng MH. Fibrin sealant promotes migration and proliferation of human articular chondrocytes: possible involvement of thrombin and protease-activated receptors. Int J Mol Med. 2006; 17(4): 551-8.

x PUBLICATIONS – ASSOCIATED TO THE THESIS TOPIC

(1) Jones CW, Smolinski D, Willers C, Yates P, Keogh A, Fick D, Kirk T, Zheng MH. Laser scanning confocal arthroscopy of a fresh cadaveric knee joint. Osteoarthritis and Cartilage; July 5, 2007.

(2) Chen JM, Willers C, Xu J, Wang A, Zheng MH. Autologous Tenocyte Therapy Using Porcine-Derived Bioscaffolds for Massive Rotator Cuff Defect in Rabbits. Tissue Eng. 2007; 13(7).

(3) Stoffel K, Willers C, Korshid O, Kuster M. Patellofemoral contact pressure following high tibial osteotomy: a cadaveric study. Knee Surg Sports Traumatol Arthrosc. 2007.

(4) Lin Z, Willers C, Xu J, Zheng MH. The chondrocyte: biology and clinical application. Tissue Eng. 2006; 12(7):1971-84.

(5) Wood D, Robertson B, Willers C, Zheng MH. Matrix-induced Autologous Chondrocyte Implantation (MACI®): surgical procedures and postoperative care in the Australian experience. IN: Zanasi S, Brittberg M, Marcacci: Basic Science, Clinical Repair and Reconstruction of Articular Cartilage Defects: Current Status and Prospects. Timeo Editore Press 2006.

(6) Zheng MH, Willers C, Wood D. Matrix-induced Autologous Chondrocyte Implantation (MACI®): biological and clinical evaluation. IN: Zanasi S, Brittberg M, Marcacci: Basic Science, Clinical Repair and Reconstruction of Articular Cartilage Defects: Current Status and Prospects. Timeo Editore Press 2006.

xi ABBREVIATIONS

µL Microlitres µm Micrometers µM Micromolar 6MWT Six minute walk test AB Alcian Blue AC Autologous chondrocytes ACI Autologous chondrocyte implantation ACI-C Collagen-covered autologous chondrocyte implantation ACI-P Periosteal autologous chondrocyte implantation ACL Anterior cruciate ligament ADL Activities of daily living ANOVA Analysis of variance BMI Body mass index BMP Bone morphogenetic protein Bp base pairs C Celsius Ca Calcium CaCl2 Calcium chloride cDNA Complementary deoxyribonucleic acid CFI Carbon fibre implantation CI Confidence interval CLSM Confocal laser scanning microscope CM Collagen membrane cm Centimetre coll Collagen CPM Continuous passive motion DAB Diaminobenzidine DDW Double distilled water dNTP Deoxynucleotides triphosphate DVT Deep vein thrombosis ELISA Enzyme-linked immunosorbent assay FBS Fetal bovine serum FC Fibrocartilage xii FDA Food and drugs administration Fig. Figure FOV Field of view FS Fibrin sealant GAG Glycosaminoglycan GAPDH Glyceraldehyde 3-Phosphate Dehydrogenase H&E Haematoxylin and Eosin

H2O2 Hydrogen peroxide HA Hyaluronic acid HLC Hyaline-like cartilage HTO High tibial osteotomy ICRS International cartilage repair society IgG Immunoglobulin G IKDC International knee documentation committee KCl Potassium chloride KOOS Knee injury and osteoarthritis outcome score KQOL Knee-related quality of life LSCA Laser scanning confocal arthroscope M Molar MACI Matrix-induced autologous chondrocyte implantation MDSC Muscle derived stem cell MFC Medial femoral condyle MFX Microfracture

MgCl2 Magnesium chloride min minutes ml Millilitre mm Millimetre MMP matrix metalloproteinase MOCART Magnetic resonance observation of cartilage repair tissue MRI Magnetic resonance imaging mRNA Messenger ribonucleic acid MSC Mesenchymal stem cell MUA Manipulation under anaesthesia MW Molecular weight N Newton xiii n Number NaCl Sodium chloride OA Osteoarthritis OATS Osteochondral autograft transfer system ORA Overall repair assessment PAR Protease activated receptor PBS Phosphate buffered saline PCL Posterior cruciate ligament PD Proton density PGA Polyglycolic acid Post-op Post-operative RNA Ribonucleic acid ROLB Removal of loose bodies ROM Range of motion RT Room temperature RT-PCR Reverse transcriptase polymerase chain reaction SD Standard deviation sec Second (s) SEM Standard error of mean SEM Scanning electron microscopy SF-36 Short form 36 Sig. Significance TBS Tris buffered saline TEM Transmission electron microscopy TGA Therapeutic goods administration TTT Tibial tuberosity transfer U Units UV Ultraviolet WAZ Willers and Zheng

xiv LIST OF FIGURES

PAGE CHAPTER 2: Figure 1: Massive full-thickness cartilage defect in the femoral trochlea 12 covering nearly the entire articular surface.

Figure 2: Histology of poor mosaicplasty repair integration at 1 year 15 follow-up. A clear space is visible between the transplanted osteochondral plug (right) and the native articular cartilage (left). The transplanted plug also appears to have deteriorated, with fibrillated edges and a predominantly fibrocartilaginous matrix morphology.

Figure 3: Medial femoral condyle defect treated by collagen-covered 17 autologous chondrocyte implantation (CACI). Microsutures can be seen radiating from the periphery of the defect, with fibrin sealant seen being added to seal the previously injected chondrocyte solution.

Figure 4: Large cartilage defect in the femoral head treated by Matrix- 18 induced Autologous Chondrocyte Implantation (MACI®). As illustrated, the MACI product allows effective shaping of the graft to the defect site, without the need for microsuture.

Figure 5: Autologous human chondrocytes aboard a type I/III collagen 18 membrane after 3 days culture. Cells readopt their differentiated spherical morphology after monolayer culture and carpet the scaffold congruously with the underlying collagen matrix.

Figure 6: Hyaline-like cartilage repair, with seamless integration 19 between the repair and native cartilage (centre). Image acquired from a matrix-induced autologous chondrocyte implantation patient at 12 months postoperatively.

xv PAGE Figure 7: The 3 stages of tissue regeneration by cell-based tissue 23 engineering products. The outcome prognosis of cell-based therapeutic products is based on the combined success of both product manufacturing, surgeon implantation of the product, and maturation of the product postoperatively through appropriate patient rehabilitation.

CHAPTER 3:

Figure 1: (A) Histology of chondrocyte migration from collagen 41 membrane to FS from 12 to 48 hrs. Chondrocyte migration from the collagen membrane towards the fibrin surface was seen after 12 hrs culture (arrows). At 24 and 48 hrs, aggregation of chondrocytes on the surface of the FS is clearly evident. In contrast, chondrocyte-seeded collagen membrane alone at 12, 24, and 48 hrs showed in-growth of chondrocytes into the collagen matrix of the membrane. All sections were stained with H&E. Magnification=25X. (B) Chondrocyte migration and in-growth into FS at 15 days under phase-contrast microscopy. As seen at both 25X and 40 X magnifications, chondrocyte in-growth was visualized within the FS (arrows). AC: autologous chondrocytes, CM: collagen membrane, FS: fibrin sealant.

Figure 2: Proliferative response of human articular chondrocytes to 41 different concentrations of thrombin. Chondrocytes were seeded onto 96- well plates, serum-starved for 24 hrs and then stimulated with varying concentrations of thrombin (0.1-10 U/ml) for an additional 24-48 hrs.

Vehicle (CaCl2, 40mM) served as a control. BrdU-incorporation as a parameter of proliferation was measured, and expressed as absorbance at 450 nm. Data are expressed as mean ± SEM (n=6). *p<0.005; **p<0.0005.

Figure 3: mRNA expression of PARs in human articular chondrocytes. 41 Total RNA was isolated from cultured primary human chondrocytes and subjected to RT-PCR using PAR isoform specific primers. Reaction products were resolved by agarose gel (1.5%) electrophoresis, stained with ethidium bromide, and visualized under a UV transilluminator. PAR-1 xvi (708 bp) and PAR-3 (382 bp) are highly expressed in chondrocytes PAGE whereas PAR-2 (582 bp) and PAR-4 (392 bp) exhibit moderate to weak expression respectively. β-actin (240 bp) served as an internal loading control.

Figure 4: Immunocytochemical detection of PARs in cultured human 42 chondrocytes. Strong expression of both PAR-1 and PAR-3 is evident in human chondrocytes. PAR-2 displayed moderate chondrocytic expression whilst negligible PAR-4 staining was detectable. Immunostaining for S- 100 served as a positive marker for chondrocytes. Sections were viewed at 250X and 400X magnification.

Figure 5: Subcellular localisation of PARs in human chondrocytes. 42 Chondrocytes were grown on 8-well chamber slides, fixed with 4% paraformaldehyde, and then immunostained for PAR isoforms (green). All slides were countered stained with Hoescht 33342 to visualise nuclei (blue) and images were recorded by confocal microscopy. Marked staining of the plasma membrane and cytosol was detected for PAR-1,-3 and -4. PAR-2 also shows plasma membrane staining and as well as a sub-population of perinuclear vesicles (inset). Bar=10 μm

Figure 6: Intracellular calcium responses in human chondrocytes elicited 42 by thrombin and PAR agonist peptides. Human chondrocytes were loaded with Fura-2/AM and incubated either with thrombin (1 U/ml) or PAR-1, 3, 4 agonist peptides at 400 µM. Thrombin (A) and PAR-1 agonist peptide SFLLRN-NH (B) induced steep intracellular calcium elevations upon 2 addition, but thrombin elevation did not return to baseline. All responses were measured over a 5 min period and results were expressed as a fluorescence ratio (340/380 nm). Traces are representative of at least 5 independent experiments.

xvii PAGE CHAPTER 4: Figure 1: (A) Macroscopic view of MACI-treated condyle. MACI graft 56 was completely intact across entire defect with full-thickness infill achieved. Cartilage appeared pink in comparison to surrounding tissue. Area surrounding and immediately adjacent to MACI-treated region appears normal AC with no surface fibrillations or evidence of degenerative changes. (B) LSCA image of MACI repair tissue. Brightly stained round chondrocytes (C) persisted in high densities in the repair tissue. (C) Hematoxylin and eosin histology (original magnification, X100) of MACI-treated condyle with 100% tissue infill, excellent integration, and surface continuity. Note native AC (N) to repair tissue interface (MACI). Repair tissue was hyalinelike cartilage. (D) MRI image of MACI treatment (arrow indicates medial femoral condyle). Good integration of the graft was seen with smooth surface features, homogeneous consistency, and no edema.

Figure 2: Modality scores for each assessment modality at 8, 10, and 12 57 weeks after surgery. Data expressed as mean _ SD. Significant relationships (multiple pairwise Student t-test, p<0.05) expressed for MACI versus other treatments. (A): Condyle ICRS modality scores showed a high degree of variability between treatment groups and time points. Condyle MACI repairs were significantly superior to controls at 10 weeks (p<0.05). (B) Trochlea ICRS modality scores demonstrated the general superiority of MACI repairs, with significant superiority at 10 weeks (p<0.05). (C) LSCA modality scores for condyle treatment groups showed neither MACI nor collagen-only scaffold groups demonstrated significant improvement in cartilage regeneration (p<0.05). (D) LSCA modality scores for trochlea showed MACI was superior to the collagen- only scaffold treatment group at 10 weeks (p<0.05). (E) Histological modality scores for condyle treatment groups showed neither MACI nor collagen-only scaffold treatments demonstrated significant improvement in cartilage regeneration (p<0.05). (F) Histological modality scores for trochlea treatment groups showed MACI evidenced a significant improvement compared to untreated controls at both 8 and 10 weeks xviii (p<0.05). (G) MRI modality scores showed improvement for MACI PAGE condylar regeneration compared to collagen only and untreated controls at all three time points, with significance at 10 weeks (p<0.05).

Figure 3: Representative LSCA images with the corresponding modality 58 score component breakdown provided in Supplementary Material (Table S1). (A) LSCA image of native articular cartilage demonstrating brightly stained classically paired chondrocytes (PC). LSCA score: 17. ORA grade I: normal. (B) LSCA image of MACI repair tissue from 10-week trochlear repair. High density round chondrocytes (C) persist in repair tissue. LSCA score: 15.ORAgrade II: nearly normal. (C) LSCA image of degenerative tissue following creation of untreated control defect 10-week trochlear repair. A low density spindly shaped infiltrate (S) is evident deep in a necrotic lesion among whirls of connective tissue (W). LSCA score: 10. ORA grade III: abnormal. (D) LSCA image of fibrous tissue following spontaneous repair of untreated control defect 10-week trochlear repair. High densities of small spindle shaped fibroblasts (F) are evident. LSCA score: 7. ORA grade IV: severely abnormal.

Figure 4: Overall treatment effect independent of time point for (A) 59 trochlea and (B) condyle treatment sites. A significant improvement in mean modality score was seen across all assessment modalities for MACI treatment compared to collagen-only bioscaffold or control (except condyle histology). Kruskal–Wallis Asymp Sig values are presented. Data expressed as mean ± SD.

Figure 5: Lagrange stress versus relative vertical displacement. MACI 60 represented by circles, control (fibrous) tissue by triangles, and native tissue by squares.

xix PAGE CHAPTER 5: Figure 1A–F: Morphological analysis of chondrocyte seeding onto ACI- 71 Maix™ bilayer collagen scaffold by electron microscopy. (A) Cross- sectional SEM imaging of the scaffold shows the differential organization of the collagen matrix. (B) SEM imaging of the cell-occlusive compact arrangement of collagen in the smooth surface. (C) The rough surface of the collagen scaffold showing its loose collagen matrix within the scaffold under SEM. (D) SEM imaging of the chondrocyte seeded scaffold shows the differentiated globular chondrocyte appearance and their attachment to the collagen fibers of the scaffold. (E) High magnification SEM of an individual chondrocyte seeded onto the collagen scaffold shows cell attachment via cytoplasmic (philopodia) projections (arrow) after seeding. (F) TEM imaging of chondrocyte attachment with the scaffold showing the presence of cytoplasmic projections anchoring cells to the collagen fibres of the scaffold (15,000X magnification).

Figure 2A-D: Phenotypic analysis of chondrocytes following seeding onto 71 ACI-Maix™ bilayer collagen scaffold. (A) S-100 positive staining shows that the integrated cells within the collagen matrix are chondrolineage, and shows their distribution throughout the rough surface of the scaffold (original magnification, X20). (B) Staining of the chondrocyte seeded scaffold for Type II collagen was positive, which suggests chondrolineage cell presence and indicates that the synthesis of collagenous matrix in the seeded scaffold is active prior to implantation (original magnification, X20). (C) Analysis of 30 patients for percentage S-100 and collagen II positivity evidenced that 80% of patients had 80% or greater cells positive for S-100 protein, whilst 50% of patients had 80% or greater cells positive for type II collagen. Interestingly, 13% of assessed patients showed no evidence of type II collagen upon staining. (D) RT-PCR comparison shows that human chondrocyte aggrecan and type II collagen gene expression of the same patient is maintained from monolayer culture through to cell inoculation onto the collagen scaffold. Aggrecan and Type II collagen expression seemed slightly higher in the seeded chondrocytes compared to monolayer culture. Lane 1 represents molecular weight standard bands; xx Lanes 2, 4, and 6 represent chondrocytes cultured in monolayer conditions; PAGE and Lanes 3, 5, and 7 represent chondrocytes seeded onto ACI-Maix™ scaffold.

Figure 3A-D: Histology of MACI® induced cartilage repair from 48 hours 72 to 6 months postoperatively. (A) Histology micrograph of the transitional zone of the regenerated tissue at 48 hours shows the spherical morphology of implanted autologous chondrocytes within the fibrin glue. (Stain, hematoxylin and eosin; original magnification, X100). (B) The 21-day repair photomicrograph shows a heterogeneous mix of spherical chondrocytes and cartilage-like matrix within mesenchymal tissue. Fibrin glue is shown within the transitional zone of the regenerated tissue at 21 days (Stain, hematoxylin and eosin; original magnification, X100). (C-D) Histology of hyaline-like regeneration is shown at 6 months postoperatively. (C) Hyaline-like cartilage was seen in the superficial zone of the regenerated cartilage with spindle shaped chondrocytes, mature chondrocytes within their hyalinelike matrix lacunae, and zonal organization similar to a healthy state (Stain, hematoxylin and eosin; original magnification, X200). (D) Type II collagen immunohistochemistry 6 months after collagen scaffold autologous chondrocyte implantation demonstrated strong positive staining in the radial zone reminiscent of native tissue (original magnification, X200).

Figure 4A–C: Histology micrographs from 8 months postoperatively show 72 the predominance of hyalinelike regeneration. (A) Low magnification of the biopsy tissue shows the hyaline-like appearance of the regenerative cartilage, with spindle shaped cells in the superficial zone and rounded cells in the transitional and radial zones (Stain, hematoxylin and eosin; original magnification, X20). (B) Globular chondrocytes within their matrix lacunae are seen in the transitional zone of the regenerated cartilage, illustrating the maturation of tissue architecture. Collagen fibrils (arrow head), possibly derived from the scaffold (Stain, hematoxylin and eosin; original magnification, X200). (C) Some residual fibrin sealant was observed in the transitional zone of the regenerative cartilage (Stain, hematoxylin and eosin; original magnification, X200).

xxi PAGE Figure 5A-D: Histology of MACI® induced cartilage repair from 12 to 24 73 months postoperatively. (A) Biopsy histology micrograph from 12 months postoperatively shows the persistence of hyalinelike cartilage in the regenerated cartilage. Higher magnification of the superficial zone exhibits the smooth surface and superficial spindle cellular morphology of the regenerated hyalinelike cartilage (Stain, hematoxylin and eosin; original magnification, X100). (B) Type II collagen immunohistochemistry is shown 12 months after collagen scaffold autologous chondrocyte implantation. Note the strong positive staining in the radial zone reminiscent of native tissue (original magnification, X100). (C) A histology micrograph from 24 months postoperatively shows the persistence of hyalinelike cartilage in the regenerated cartilage. Hyalinelike morphology of the regenerated cartilage is seen in the superficial and transitional zones, with moderate cell density, spherical chondrocytes within lacunae, and no obvious cellular architecture (Stain, hematoxylin and eosin; original magnification, X100). (D) Histology of the healthy- regenerative tissue interface 24 months after MACI®. Histology between the healthy (H) and regenerative (R) cartilage at 24 months evidenced a cleft at the defect interface. The regenerative cartilage showed higher cell density than the abutting host cartilage (Stain, haematoxylin and eosin; original magnification, X40).

Figure 6A-E: Histologic and appearance of a retrieved MACI® graft after 73 18 months repair. (A) The MACI® grafted defect appeared white and smooth in texture macroscopically, as seen with healthy cartilage, whereas the surrounding knee surface was yellow and appeared fibrillated as with osteoarthritic cartilage (photo taken after slicing for sectioning). (B) Osteoarthritis of the cartilage immediately adjacent to the MACI® grafted defect (imaged from centre of defect – dashed line in panel A), with characteristic fibrillation of the articular cartilage surface, reduced cell density and chondrocyte clustering (Stain, haematoxylin and eosin; original magnification, 40X). (C) Low magnification of the MACI® regenerate (dashed line in panel A) shows the homogeneous matrix appearance of the site and complete tissue integration (Stain, haematoxylin and eosin; xxii original magnification, 5X). (D) The MACI® grafted area showed good PAGE restoration of osteochondral architecture and regenerative tissue generally characterized as hyaline-like cartilage (Stain, haematoxylin and eosin; original magnification, 200X). (E) The hyaline-like cartilage regeneration was also rich in proteoglycan (Stain, Alcian blue; original magnification, 200X).

CHAPTER 6: Figure 1: Repair tissue composition for revised and clinically failed cases. 94 Whilst small differences were observed between the composition of revised and failed cases, none were found to be significant (P>0.05). BARS (top two rows): Light grey = weak positive; Dark grey = medium positive; Black = strong positive; Dotted white = total positivity. BARS (MMP13, Sox9, Ki67): Black = total nuclei positive. Sample size: n=12 for each antibody. Error bars indicate standard error (SEM).

Figure 2: Representative immunohistochemical images of the hyaline-like 95 cartilage (HLC) cohort. HLC was characterized (H&E) by a sparse but seemingly uniform distribution of mature chondrocytes within their lacunae, surrounded by hyaline-like matrix. Aggrecan, collagen II, collagen III, collagen VI, collagen IX, Sox-9 were generally positive. Aggrecan and all collagens were diffusely stained within the matrix of the positive tissues. Notably, collagen VI was not localised to the pericellular region of the positive biopsies. While the chondrogenic transcription factor Sox-9 was isolated to the nucleus. Collagen I stained weakly in only a few cases. Ki-67 and MMP-13 were negative in general. Magnifications – 50X: coll I; 100X: agg, coll II, coll III, coll IX, Ki-67 and MMP-13; 200X: coll VI and Sox-9.

Figure 3: Representative immunohistochemical images of the 96 fibrocartilage (FC) repair cohort. FC was characterized (H&E) by an irregular distribution of predominantly spindle-shaped chondrocytes within a dense irregular fibrous matrix. Collagen III, collagen VI, collagen IX, xxiii and Ki-67 were generally positive. Collagens III, VI, and IX were diffusely PAGE stained within the matrix of the positive tissues, while the proliferative marker Ki-67 was isolated to the nucleus (no counterstain). Whereas aggrecan, collagen I, collagen II, Sox-9 and MMP-13 were negative in general. Magnifications - 100X: agg, coll I, coll II, coll III, coll VI, coll IX and MMP-13; 200X: Sox-9 and Ki-67.

Figure 4: Tissue composition for hyaline-like, fibrocartilage, and mixed 97 repair cases. Collagen II production was significantly (P<0.05) increased in the both the HLC and mixed repairs compared to FC repairs. Collagen IX production was significantly (P<0.05) increased in HLC repairs compared to FC repairs. While a significant (P<0.05) increase in the proliferative marker Ki67 was observed in the FC repairs. BARS (top two rows): Light grey = weak positive; Dark grey = medium positive; Black = strong positive; Dotted white = total positivity. BARS (MMP13, Sox9, Ki67): Black = total nuclei positive. Sample size: n=12 for each antibody. Error bars indicate standard error (SEM).

Figure 5: Stiffness testing and associated histology of a complete cartilage 98 repair site on the medial femoral condyle. (A) The superior portion of the repair site (M1) displayed a maximum stiffness of 4.32N, which was 55% that of the control adjacent tissue (C1) at 7.73N. (B) This reduced stiffness correlated to a lack of proteoglycan in M1 compared to C1. (A) In contrast, the inferior portion of the repair site (M2) displayed a maximal stiffness of 6.64N, which was 85% that of the adjacent control tissue (C2) at 7.81N. (B) This improved stiffness appears to be correlated to an observable increase in proteoglycan content in M2 compared to M1. All images shown are Alcian Blue staining. Magnification – 50X for all images.

xxiv CHAPTER 7: PAGE Figure 1: Paradigm of matrix-induced autologous chondrocyte 127 implantation (MACI) cartilage regeneration. (1) Implantation of chondrocyte seeded membrane into the fibrin sealant-covered base of the debrided chondral defect (day of implantation). (2) Cell migration of chondrocytes from the cambium surface of the membrane into the fibrin sealant matrix. Host resorption of the collagen membrane has also commenced (2-5 days following implantation). (3) Matrix production by implanted autologous chondrocytes. Type II collagen, aggrecan and other matrix proteins important for healthy articular cartilage function are synthesised by the newly implanted cells (1-12 months following implantation). (4) Matrix maturation and hyaline-like/hyaline cartilage formation. Cartilage infill is complete, chondrocyte morphology and surrounding matrix appears healthy (or similar to surrounding native tissue) and graft cartilage is well integrated with the adjacent cartilage (12- 24 months following implantation).

Figure 2: The graduated return to weightbearing administered to patients 128 during functional rehabilitation following their MACI surgery. Gradual loading of the joint is conducted to stimulate maturation and adaptation of hyaline-like cartilage infill through physiologically induced chondrocyte biosynthesis.

Figure 3: Improvement in six-minute walk distance to 24 months 129 postoperatively (n = 28). An increased six-minute walk distance was noted from 6 months postoperatively compared to preoperatively. Scores for this parameter were suppressed at 3 months, and and improvement occurred predominantly in the first 12 months onwards (P<0.05).

Figure 4: Improvement in the five sub domains of KOOS to 24 months 130 postoperatively (n = 28). Observed improvements in knee pain, symptoms, and ADL were noted predominantly in the first 12 postoperative months onwards. While improvements in sport and recreation function increased linearly from 3 months onwards, and knee related quality of life improved significantly from 3 months onwards (P>0.05). Improvement in the sport and recreation and knee related quality of life subscales were reduced xxv compared to the other subscales due to the higher functional demand of PAGE these components. Total KOOS scores (0 = extreme knee problems and 100 = no knee problems), ADL = activities of daily living, Sport&Rec = sport and recreation function, KQOL = knee-related quality of life.

Figure 5: Improvement in MRI composite score to 24 months 131 postoperatively. MACI patients demonstrated an increased MRI composite score that improved significantly (P<0.001) from 3 to 24 months postoperatively, with predominant increase seen in the first 12 months. Specifically, at 12 months, good to excellent filling of the defect had increased to 76% of grafts (from 45%), and signal intensity had improved to 93% of grafts (from 28%). * P<0.001 compared to 3 months.

Figure 6: Sagittal proton density fast spin echo magnetic resonance image 132 of a MACI graft in a patient treated for a full thickness chondral defect of the medial femoral condyle. Compared to preoperatively (A), the MACI graft was hyperintense and of reduced thickness compared with the adjacent normal articular cartilage three months postoperatively (B). At one year postoperatively (C) the MACI graft displayed a heterogeneous appearance with similar thickness to the adjacent normal cartilage. Reconstitution of the subchondral bone plate (arrows) improved markedly from the 3 month time point. At two years postoperatively, the MACI graft remained intact and demonstrated a heterogeneous graft signal compared to the adjacent native cartilage. Border integration was smooth, with no radiographical evidence of fissures or clefts between the native cartilage or within the graft, and restoration of the subchondral plate appeared almost complete.

xxvi LIST OF TABLES

PAGE CHAPTER 2: Table 1: Autologous chondrocyte implantation (ACI) products: Countries, 16 companies, number of peer-reviewed references, and maximum follow-up.

CHAPTER 4: Table 1. Modality scoring and overall repair assessment grading 53

Table 2. Pearson correlations for condyle treatment groups (ICRS vs. 54 WAZ vs. LSCA vs. MRI), demonstrating significant correlation between LSCA and MRI and between ICRS and MRI (p<0.05).

CHAPTER 5: Table 1: Biopsied case information. 69

Table 2: Summary of histological findings of biopsied cases. 74

Table 3: Comparison of hyaline-like cartilage regeneration between 74 investigators

CHAPTER 6: Table 1: Antibody list with associated antigen retrieval and dilution factor 99 used.

Table 2: Repair type distribution for revised and clinically failed repair 100 biopsies.

Table 3: Repair type distribution and clinical status by anatomical location 101 of defect.

Table 4: Demographics, implant details, and repair compositions for 102 revised and clinically failed biopsies. xxvii PAGE CHAPTER 7: Table 1: Surgical histories for the MACI cohort. Procedures conducted 133 before, during and after matrix-induced autologous chondrocyte implantation (MACI).

Table 2: Statistics summary for the six-minute walk test, and the five sub- 134 scales of the KOOS score (pain, symptoms, activities of daily living, sport and recreation function, and knee related quality of life).

CHAPTER 8: Table 1: Ten components of the patient satisfaction questionnaire. 155

Table 2: Patient procedures before, during, and after MACI® (type and 156 percentage stated).

Table 3: Raw questionnaire outcomes 157

Table 4: Statistical analysis of demographic, pre- and post-operative 158 variables with overall satisfaction scores.

xxviii THESIS ABSTRACT

Articular cartilage has no vascular, neural, or lymphatic supply, and hence no intrinsic capacity to self-repair following injury. These physiological limitations, combined with the inability of local chondrocytes to contribute to the repair process, translate to poor structural and functional outcomes in these troublesome defects, and osteoarthritic deterioration with time. Subsequently, many surgical therapies have been trialed to stimulate cartilage repair, but none have produced reliable outcomes. Hence, cartilage repair research has been broadened, with many investigators now focused on cell-based treatment. Smith began a revolution of autologous cell research when in 1965 she isolated chondrocytes from articular cartilage and transplanted them into fresh cartilage nodules (Smith, 1965). Since, new technologies and improved techniques have seen autologous chondrocyte implantation (ACI) widely accepted for use in clinical orthopaedics (Bentley et al., 2003; Brittberg et al., 1994; Grande et al., 1989; Peterson et al., 2002). At present, matrix-induced autologous chondrocyte implantation (MACI) is the most surgically simple form of ACI, boasting clinical outcomes comparable to any technique on the market, and far less complications compared to the first generation of ACI - periosteal ACI (Bartlett et al., 2005; Behrens et al., 2006; Gigante et al., 2006; Henderson et al., 2004; Marlovits et al., 2005; Minas, 2001; Willers et al., 2007; Zheng et al., 2007). But whilst MACI has been adopted by the orthopaedic surgeon for articular cartilage repair, many of the molecular, histological, and clinical factors governing patient outcomes are still largely understudied.

Firstly we assessed the bioactivity of fibrin sealant (FS - Tisseel®), a critical component of MACI, on the migration and proliferation of human articular chondrocytes in vitro. We also looked to elucidate the associated molecular mechanisms of thrombin, a key active ingredient in FS, by examining the expression and activation of protease- activated receptors (PARs), established thrombin receptors. All four PAR isoforms were detected in human chondrocytes, with PAR-1 being the major isoform expressed. Moreover, thrombin and PAR-1, but not other PAR-isoform-specific peptide agonists, were found to induce rapid intracellular Ca2+ responses in human chondrocytes in calcium mobilization assays. Together, these data demonstrate that FS supports both the migration and proliferation of human chondrocytes. We propose that these effects are mediated, at least in part, via thrombin induced PAR-1 signaling in human chondrocytes. xxix Following on from our previous rabbit MACI model, we established an ovine model of articular cartilage repair and examined the ability of the non-destructive laser-scanning confocal arthroscope (LSCA) for assessing MACI repair (Willers et al, 2005). Repair outcomes were examined using LSCA, magnetic resonance imaging, histology, macroscopic ICRS grading, and biomechanical analysis. Pearson correlation analysis demonstrated the correlation between LSCA, MRI, and ICRS grading. Moreover, testing of overall treatment effect independent of time point revealed significant differences between MACI and control groups for all sites and assessment modalities, except condyle histology. Biomechanical analysis suggests that while MACI tissue may histologically resemble native tissue in the early stages of remodeling, the biomechanical properties remain inferior, at least in the short term. Hence, we demonstrated the potential of a multisite sheep model of articular cartilage defect repair and its assessment via non-destructive LSCA.

The thesis also sought to better our understanding of the histological paradigm of tissue repair after MACI. From a cohort of 56 MACI patients, we examined the phenotype of chondrocytes seeded on type I/III collagen scaffold, and conducted progressive histologic assessment over a 6 months period. Coincidental cartilage biopsies were obtained at 48 hours, 21 days, 6months, 8months, 12months, 18months, and 24months. Our data showed that chondrocytes on the collagen scaffold appeared spherical, well integrated into the matrix, and maintained the chondrocyte phenotype as evidenced by aggrecan, type II collagen, and S-100 expression. Progressive histologic evaluation of the biopsies showed the formation of cartilage-like tissue as early as 21 days, and 75% hyaline-like cartilage regeneration after 6 months. This preliminary assessment of MACI repair tissue maturation advocates the cell-based therapy as an alternative to the surgical treatment of cartilage injury, showing hyaline-like cartilage as early as 6 months postoperatively.

However, not all MACI repairs are successful, with some cases needing revision or deteriorating to be classified as clinical failure. To better appreciate the biology of these cases, we investigated the histological and immunohistochemical characteristics of revised and failed MACI repair tissues. We examined the matrix profiles of repair biopsies taken from revised and clinically failed MACI cases by quantitative immunohistochemical study using antibodies specific to aggrecan, collagens I, II, III, VI, and IX, Sox-9, Ki-67 and MMP-13. We also stiffness tested an intact clinically xxx failed repair site. Histologically, the majority of these biopsies (n=39) were hyaline-like (HLC) and fibrocartilage (FC) in both the revised (30% and 38% respectively) and failed (34% and 22% respectively) cases. Compositionally, more revised cases were positive for aggrecan, collagens VI and IX, and Ki67 compared to failed cases, but not quantitatively different (P<0.05). More HLC biopsies were positive for aggrecan and collagen II (compared to the FC group), with diffuse and often colocalized matrix distribution. The majority of HLC biopsies stained positive for Sox-9, whereas FC cases were negative. Most (75%) FC biopsies were positive for Ki-67, compared to the HLC group with 25%. MMP-13 was negative in all biopsies. Qualitatively, reduced collagen II and IX, and increased Ki67 production was noted in FC biopsies (P<0.05). An intact repair site showed FC with 30% greater stiffness in the inferior portion compared to the superior, with an associated proteoglycan content increase. Revised and failed biopsies are predominantly hyaline-like and fibrocartilage in repair type, are histologically dissimilar to healthy cartilage, and do not differ in composition. Hyaline-like repairs show lower proliferation but improved matrix to fibrocartilage. This data furthers our knowledge into failed and revised cartilage repair following MACI, and illustrates the inferior composition of such repair tissues compared to healthy articular cartilage.

While histology is important for understanding the underlying biology of MACI, clinical improvement is inevitably the most important consideration when assessing treatment outcomes. To this end, we assessed the functional and MRI outcomes of 31 MACI patients over a 24 month follow-up period. Patients demonstrated a significant improvement in walk distance and all five KOOS subscales from 3 to 24 months after MACI surgery, with the most substantial gains in the first 12 months. Similarly, patients also demonstrated significant improved MRI scoring from 3 to 24 months, with post- hoc analysis demonstrating improvement predominantly in the first 12 months, then plateauing thereafter. A 10% incidence of hypertrophic growth following MACI was observed. Interestingly, we found that MRI score significantly correlated to all functional outcome parameters, whilst defect size and cell number showed no correlation to any functional parameters. This sub-study supports MACI as a cell-based treatment for articular cartilage injury, with significant functional improvements by one year postoperatively. While also suggesting that MRI may be used to predict functional change. Both of these findings may be used to refine the postoperative care and rehabilitation of MACI patients in order to maximize therapeutic efficacy.

xxxi Lastly, while MACI is becoming increasingly popular for cartilage repair, studies such as that detailed in the previous paragraph, generally involve small cohorts assessed by varied scoring systems. This makes comparisons of study data in the literature next to impossible. Therefore, we conducted a retrospective multi-centre cohort study of 202 patients using a single ten-question questionnaire covering patient symptoms, function, quality of life, and satisfaction (based on the Lysholm and Cincinnati scales). The mean follow-up was 22.7±8.6 months, mean age 36.9±10.7 years, and mean size of defect was 4.8cm2. Fifty-nine percent of patients had previous surgery to their knee. Overall results showed that 167 (83%) of the surveyed patients had a good to excellent MACI outcome. We have also found significant improvement in younger patients (<30 vs. 30- 50), those surveyed after more than 24 months (>24 months vs. 12-24 months), those participating in formalised rehabilitation (with vs. without), and those with defects located on the femoral components of the patellofemoral joint (trochlea vs. patella, medial femoral condyle vs, patella). MACI showed a high rate of good/excellent patient satisfaction and clinical outcome, with better results in younger, more active knee joints. This study suggests that MACI produces good to excellent outcomes in satisfied patients, but may be improved by postoperative conditioning of the joint.

This thesis has demonstrated biological, histological, and clinical features of the MACI technique. Our in vitro has supported the use of fibrin sealant and collagen membrane as the major material components of MACI, illustrating improved chondrocyte proliferation, migration, and chondrogenic differentiation. We have evidenced that MACI stimulates successful production of hyaline-like cartilage by 6 months, while also showing that revised and clinically failed repair tissues are predominantly hyaline-like and fibrocartilage with inferior composition. Clinically, we have documented significant improvements in patient repair structure, function, symptoms, quality of life, and satisfaction, whilst concurrently confirming sentiment within the literature regarding the importance of exercise/ rehabilitation for maximising MACI outcome. In summary, the findings presented in this thesis suggest that MACI is a biologically sound and clinically efficacious cell-based treatment option for repairing articular cartilage defects.

Chapter 1

General Introduction & Thesis Objectives

Page 1

1.1 GENERAL INTRODUCTION

As the articular cartilage covering our joints has no intrinsic capability to self-repair following injury, patient symptoms degenerate with time and leave troublesome cartilage defects with poor clinical prognosis. Of all joints, the knee is the most commonly affected, accounting for about 75% of all lesions (Clanton and DeLee, 1982;

Obedian and Grelsamer, 1997). Additionally, 63% of a large cohort of knee arthroscopies (31,516 patients) has revealed chondral lesions and 20% full-thickness

(grade IV) lesions, with 2.7 lesions per knee on average, mainly to the medial femoral condyle (Curl et al., 1997). Injury is predominantly diagnosed in young physically active males, with two to three times more prevalence in men than in women (Clanton and DeLee, 1982). Etiologically, the majority of lesions are subsequent to a traumatic event (Boden et al., 1997; Nomura et al., 2003). Indeed some authors argue that articular cartilage injury can be classified as either those with a history of trauma, or those without (Birk and DeLee, 2001). Long-term, symptomatic articular cartilage lesions also show a strong prognostic correlation to osteoarthritis (Bentley and Minas,

2000). It is estimated that approximately 60% of patients will have significant symptomatic OA within 20 years of generating an articular cartilage defect. With an aging population, OA will represent an increasingly significant healthcare burden. In the

United States in 1997, $186.9 billion (77% of total musculoskeletal costs) was attributed to cartilage degeneration (Yelin, 2003). Globally, the costs of musculoskeletal illness have been shown to have risen, accounting for up to 2.5% of the gross national product for countries including the USA, Canada, UK, France and Australia (March and

Bachmeier, 1997).

Whilst there have been substantial developments in the treatment of articular cartilage injury over recent decades, there is still no documented treatment technique which can Page 2 offer reliable restoration of the unique structural and functional characteristics that define healthy hyaline articular cartilage. The myriad of conventional surgical therapies trialed for cartilage repair have all been limited by inferior repair tissue, variable functional outcomes, and elevated rates of postoperative complication. These techniques include arthroscopic debridement and lavage, microfracture, and osteochondral autografting (mosaicplasty).

Subsequently, the treatment of cartilage repair has moved away from surgical interventions, with research now heavily focused on cell-based approaches. Specifically, the most promising candidate at present for qualitative cartilage repair is autologous chondrocyte implantation (ACI). In short, ACI involves two surgeries. Firstly, usually via arthroscopic surgery, a small amount of cartilage is harvested from a nonweightbearing region of the joint. This biopsy is then transferred to a tissue culture laboratory where chondrocytes are isolated and cultivated to gain enough cells for implantation (approx. 10 million cells). Secondly, usually via open arthrotomy, the defect is debrided to remove any native tissue repair, and the chondrocytes are implanted as either a cell suspension or cell-bioscaffold construct. However, whilst the first generation of ACI (periosteal ACI, ACI-P) pioneered the therapeutic potential of this technology, exhibiting a high percentage of improved functional outcomes and hyaline-like cartilage repair, many complications have overshadowed its clinical success

(Brittberg et al., 1994). Problems such as periosteal hypertrophy, graft delamination, and periosteal calcification have raised questions regarding the technique’s biology, and instigated research into exogenous, non-living bioresorbable materials for a more stable integration of chondrocytes into the defect (Briggs et al., 2003; Cherubino et al., 2003;

Driesang and Hunziker, 2000; Minas and Nehrer, 1997; Ueno et al., 2001).

Page 3

Accordingly, the second generation of ACI, collagen-covered ACI (ACI-C), capitalized on the known biocompatibility of type I/III collagen membrane to address the complications associated with using periosteum (as in ACI-P). Collagen membrane has been supported by various studies for its capacity to remove the need for periosteal excision, thereby abolishing associated complications, minimizing invasiveness and operating time, and simplifying surgical technique (Briggs et al., 2003; Cherubino et al.,

2003; Nehrer et al., 1998; Russlies et al., 2002; Yao et al., 2000). The material properties of the membrane allow the chondrocyte suspension to be securely sealed within the defect, whilst its inert composition discourages hypertrophic growth. But although ACI-C has illustrated significant (P=0.01) clinical improvement at one year and 79% good-excellent repair appearance under arthroscopy, the need to suture the collagen membrane to the defect remains an undesirable component of the technique

(Bartlett et al., 2005).

Fibrin sealant, well known for its tissue adhesive properties, was then introduced as an alternative method for fixing the collagen membrane into the defects. Furthermore, fibrin sealant acts as an adhesive substrate and promotes the migration and biosynthesis of chondrocytes from the membrane (Kirilak et al., 2006). Notably, Brittberg reported that fibrin sealant (Tisseel®) inhibits chondrocyte migration into chondral defects, however numerous subsequent studies, including the authors research, contradict this lone result (Brittberg et al., 1997; Kirilak et al., 2006; Willers et al., 2005; Zheng et al.,

2007). Also, no obvious clinical complications using fibrin sealant in ACI have been reported to date. Therefore, a third generation of ACI was developed to exploit the benefits of both collagen membrane and fibrin sealant in chondrocyte-based cartilage repair.

Page 4

Matrix-induced autologous chondrocyte implantation (MACI) has become one of, if not the most popularized commercial forms of the ACI technique, with over 6000 MACI cases estimated worldwide. MACI uses a bilayer type I/III collagen membrane as a bioscaffold to allow the inoculation and integration of autologous chondrocytes, and the formation of a stable cell-scaffold construct. Upon implantation, the construct is shaped and fixed to the defect base using only fibrin tissue sealant and digital pressure, to produce a water-tight biochamber for repair tissue and cartilage maturation. Cherubino was the first to support the MACI technique, publishing improved clinical and functional outcomes, no complications, and MRI-visualized hyaline-like cartilage

(Cherubino et al., 2003). More recently, a 5-year follow-up of MACI patients has reported significant clinical improvement with no instance of hypertrophy or calcification of the graft; advocating the product as a suitable but cost-intensive therapy in cartilage repair (Behrens et al., 2006).

However, there are also several biological, histological, and clinical aspects of the

MACI technique, which have not yet received attention within the literature. In particular, the molecular and biological mechanisms underlying the histological characteristics of MACI-induced cartilage repair (both successful and failed), novel methods for assessing MACI cartilage repair, and an examination of clinical MACI outcomes and confounding factors are needed. Many of these research questions have spawned the objectives of this thesis.

Page 5

1.2 THESIS OBJECTIVES

The major objectives of this thesis were to expand our knowledge of the poorly understood molecular biology of chondrocytes within the MACI construct, especially the effect of fibrin sealant – a crucial component of MACI; examine a non-destructive imaging alternative to mechanical biopsy for assessing MACI repair tissue; evaluate the postoperative histological progression of cartilage repair tissue post-MACI; characterize the matrix composition of revised and failed MACI repair tissue; and assess MACI’s clinical capacity to restore knee structure and function, and patient satisfaction postoperatively. More specific objectives of this thesis are detailed in the following sections.

1) IN VITRO & PRECLINICAL MACI

CHAPTER 3:

The biological contribution of fibrin sealant to the MACI technique has been

questioned in the literature. Hence, the bioactive properties of Tisseel® fibrin

sealant involved in controlling chondrocyte behaviour following implantation

were analysed. Specifically, the influence of thrombin on chondrocyte migration

and proliferation, and the importance of protease activated receptors (PARs)

were examined. This was conducted to increase our understanding of the

molecular mechanisms of MACI contributing to cartilage repair.

Page 6

CHAPTER 4:

At present, destructive mechanical biopsy is the most used and accepted method

of examining cartilage repair at a macroscopic level. Due to the invasiveness of

this practice, consent to biopsy successful repairs is rare. Therefore, we

considered the use of laser-scanning confocal arthroscopy (LSCA) as a non-

destructive tool to image the cellular and matrix features of cartilage repair site,

whilst also evaluating the regenerative capability of MACI in a large animal

(sheep) model. This was conducted to both evaluate a possible non-destructive

imaging alternate to tissue biopsy, and to understand the action of MACI in a

multi-site sheep model.

2) CLINICAL MACI

CHAPTER 5:

Although it is appreciated that MACI treatment generally results in hyaline-like

cartilage repair, this may vary, and the time-frame for repair maturation is poorly

understood. Therefore, the biological progression of MACI repair histology was

evaluated from immediately after implantation to mid-term follow-up using

biopsied human MACI induced tissue at various time points of repair. The

ultrastructure and phenotype of the chondrocyte-seeded collagen membrane

construct used in MACI was also examined. This study documented the impact

of the collagen membrane on chondrocyte behaviour, and bettered our

understanding of the timeline in which repair tissue forms following MACI.

Page 7

CHAPTER 6:

Much emphasis is placed on the characterization of successful MACI, but little

research has shed light on the type and composition of cartilage repair in revised

and failed cases. Herein, using histology and immunohistochemistry, we

characterized revised and failed MACI repair tissues using biopsied cartilage.

This was conducted to provide biological insight into the tissue type and

composition of MACI grafts that become symptomatic postoperatively, or

deteriorate to be classified as failed cases in knees destined for prosthetic

replacement.

CHAPTER 7:

Numerous studies exist in the literature that look at the clinical outcomes of

periosteal ACI and collagen-covered ACI, while few have reported on those of

MACI. Accordingly, we sought to assess the functional and structural (MRI)

clinical outcomes of MACI in a small single-surgeon cohort, as well as the

relationship between these two parameters. This was conducted to expand

understanding of the restoration of the patient’s cartilage structure and function

following MACI, and gauge the timeline of patient improvement.

Page 8

CHAPTER 8:

Comparison of the outcomes achieved by the many ACI cartilage repair studies

within the literature is problematic given a lack of standardization of the

outcome measures implemented. Hence, we evaluated the clinical outcomes of a

large retrospective cohort of MACI patients using a single questionnaire

designed specifically to look at symptom relief, functional improvement, patient

satisfaction, and associated surgical variables. This was designed to gain a better

appreciation of patient improvement and product satisfaction after MACI, from

the patient’s perspective.

Chapter 2

Introduction to Cartilage Repair Techniques

Thesis publication #1: Willers C, Partsalis T, Zheng MH. Articular cartilage repair: procedures versus products. Expert Rev Med Devices. 2007; 4(3): 373-92.

Page 9

STATEMENTS OF AUTHORSHIP CONTRIBUTION FROM CO-AUTHORS

Articular cartilage repair: procedures versus products

Willers C, Partsalis T, Zheng MH

Published in the Expert Reviews in Medical Devices 2007, volume 4(3), pages 373-92.

Willers C (PhD Candidate) Major contribution to the planning, execution, analysis, and interpretation of all research. Major contribution to writing of the manuscript.

Signature of Principal Author: ……………………………… Date: .………………

Partsalis T (Research Collaborator) Minor contribution to the execution and interpretation of research. Minor contribution to writing of the manuscript.

Signature of Co-Author: ……………………..……………… Date: …………………

Zheng MH (Supervisor) Moderate contribution to planning of the research. Minor contribution to writing of the manuscript.

Signature of Co-Author: ……………………..……………… Date: …………………

Page 10

STATEMENTS OF PERMISSION FROM CO-AUTHORS

I, Theo Partsalis give permission to Craig Willers, principal author of the paper: Articular cartilage repair: procedures versus products., published in Expert Reviews in Medical Devices (2007), to include this paper as Chapter 2 of his PhD thesis entitled Matrix-induced Autologous Chondrocyte Implantation for Articular Cartilage Injury: Biology, Histology, and Clinical Outcomes.

Signature: …………………………………. Date: ………………………………….

I, Ming-Hao Zheng give permission to Craig Willers, principal author of the paper: Articular cartilage repair: procedures versus products., published in Expert Reviews in Medical Devices (2007), to include this paper as Chapter 2 of his PhD thesis entitled Matrix-induced Autologous Chondrocyte Implantation for Articular Cartilage Injury: Biology, Histology, and Clinical Outcomes.

Signature: …………………………………. Date: ………………………………….

Page 11

Review

For reprint orders, please contact [email protected]

Articular cartilage repair: procedures versus products Craig Willers, Theo Partsalis and Ming-Hao Zheng†

This review discusses the current perspectives and practices regarding the treatment of articular cartilage injury. Specifically, the authors have delineated and examined articular cartilage repair techniques as either surgical procedures or manufactured products. Although both methodologies are used to treat articular cartilage injury, there are obvious advantages and disadvantages to the application of both, with the literature providing few recommendations on the most suitable regimen for the patient and surgeon. In recent CONTENTS times, cell-based tissue engineering products, predominantly autologous chondrocyte implantation, have been the subject of much research and have become clinically Burden of cartilage injury popular. Herein, we review the most used procedures and products in cartilage repair, Native articular compare and contrast their outcomes, and evaluate the issues that must be overcome in cartilage repair order to improve patient efficacy in the future. Cartilage repair procedures Expert Rev. Med. Devices 4(3), 373–392 (2007) Cartilage repair products Comparison of efficacy Burden of cartilage injury Symptomatic articular cartilage lesions affect between cartilage repair As articular cartilage has no intrinsic capability millions of people worldwide and show a procedures & products to self-regenerate following injury, patient strong prognostic correlation to osteoarthritis Impact of concomitant symptoms generally deteriorate with time and (OA) in later life [8]. Patients with chondral joint surgery leave orthopedic surgeons with the difficult injury are at a high risk of developing OA and Expert commentary task of treating these troublesome lesions. The it is estimated that approximately 60% of Five-year view knee is the most commonly affected anatomi- patients will have significant symptomatic OA cal location, accounting for approximately within 20 years of generating an articular carti- Key issues 75% of all lesions [1,2]. Moreover, a large lage defect. As little or no data exists on the References cohort study by Curl and colleagues in the cost to society of cartilage defects alone, we Affiliations USA demonstrated that 63% of knee arthro- will focus on the cost of their pathological pro- scopies (31,516 patients) reveal chondral gression to OA. It is estimated that 68% of lesions and 20% had full-thickness (grade IV) individuals over 55 years of age have radio- lesions, with 2.7 lesions per knee on average, graphic evidence of OA [9]. With an aging mainly to the medial femoral condyle [3]. population, OA will represent an increasingly † Author for correspondence These figures are consistent with a smaller, but significant healthcare burden. In 1997, the Department of Orthopaedics, School of Pathology and Surgery, more recent, study into chondral knee total cost of musculoskeletal conditions was University of Western Australia, injury [4]. Injury is predominantly diagnosed US$240 billion (2.9% of gross domestic prod- 2nd Floor, M-block, in young physically active males, with two- to uct [GDP]) in the USA [10]. Of this figure, QEII Medical Centre, Nedlands, three-times more prevalence in men than in 77%, or $186.9 billion, was attributed to car- Perth, WA 6009, Australia women [2]. Curl and colleagues reported tilage degeneration, with indirect costs (lost Tel.: +61 089 346 3213 Fax: +61 089 346 3210 lesions in 61.6% of males but only 38.4% in wages) owing to morbidity accounting for [email protected] female patients [3]. The majority of lesions are $98.2 billion alone. Inflating these dollar val- subsequent to a traumatic insult [5,6]. Indeed, ues to those of 1996 through use of the GDP KEYWORDS: allogenic, articular cartilage some argue that cartilage injury can be classi- price, deflator yield values of $164.9 billion injury, autologous chondrocyte fied as either those who have a documented for all musculoskeletal conditions and implantation, microfracture, mosaicplasty, repair products and history of trauma or those who lack any such $71.2 billion for forms of cartilage degenera- procedures, stem cells history [7]. tion were reported in the USA [11]. The costs

Willers, Partsalis & Zheng of musculoskeletal illness have risen, accounting for up to 2.5% the subchondral plate and allow (MSCs and factors to populate of the gross national product for those countries studied, the lesion, most lesions are of partial thickness in nature and, including the USA, Canada, UK, France and Australia [12]. In hence, have a negligible cellular contribution to defect fill and 2006, it was reported that musculoskeletal illness, predomi- repair. Accordingly, many studies have demonstrated that mini- nantly OA, is still a significant socioeconomic burden in both mal chondral repair occurs initially by fibroblastic and pluri- developed and developing countries; emphasizing the need for potential cell diffusion into the defect via the subchondral bone government investment in strategies to reduce the burden of or synovial fluid, with the latter source more favored in the liter- musculoskeletal illness by education and policy reform [13]. ature [14,15]. Indeed, Hunziker illustrated that MSCs are While many treatments have been implemented over the last recruited from the synovial lining and participate in cartilage few decades to repair articular cartilage lesions, the problem repair given an appropriate defect environment for migration remains that they can be repaired but do not regenerate. The and adherence [16]. Interestingly, Dowthwaite and colleagues repair tissue that results from many of these therapeutic efforts have also recently discovered the existence of progenitor-like can range from fibrous tissue to fibrocartilage to hyaline-like cells within the superficial zone of the articular cartilage, sug- cartilage, with most repair tissues comprising a heterologous mix gesting that an appositional mechanism may also contribute to of all three tissue types. Accordingly, as tissue structure generally chondral growth and repair [17]. equates to function, the clinical outcomes seen in patients is Regardless of the origins of repair factors, the end result is unreliable and unpredictable in many cases. Arthroscopic lavage generally an incomplete infill of heterologous, fibrous repair and debridement, abrasion arthroplasty, subchondral drilling tissue (fibrocartilaginous) that lacks the biochemical character- and microfracture have been useful in removing articular debris istics necessary to withstand the daily forces distributed across and promoting the formation of fibrocartilage repair through the knee during articulation. To this end, these lesions gener- native repair mechanisms (marrow cell infiltration). However, ally present, under arthroscopic assessment, as ‘divots’ in the this type of tissue repair is functionally inferior to healthy hya- articular surface. Or, in the case of larger lesions, the surface line articular cartilage and, hence, exhibits gradual deterioration may be eburnated down to bone (FIGURE 1), with hemarthrosis with time. A possible solution to this problem may lie in the and loose bodies visible. The fibrocartilaginous attempt at self- development of tissue-engineered constructs and the utilization repair, commonly seen upon inspection of the joint, generally of cell-based tissue engineering technology to stimulate a more deteriorates over time through continued abrasion, resulting in reliable regeneration of biomechanically suitable hyaline-like to symptomatic return and, occasionally, degenerative progres- hyaline articular cartilage. sion to OA [15,18–20]. As this delicate tissue has limited self- repair, these lesions generally deteriorate under daily joint Native articular cartilage repair loading into much larger lesions, making treatment more diffi- Articular cartilage has no vasculature, lymphatics or nerve cult and involving the underlying subchondral bone in cases supply to facilitate repair following injury. The finite capability left undiagnosed for extended periods. of chondrocytes abutting the defect site to synthesize matrix for In summary, there appears to be no native regenerative repair and the inability of mesenchymal stem cell (MSC) mechanism capable of restoring the structural and biomechani- recruitment, make native defect repair in chondral lesions negli- cal competence of native hyaline articular cartilage after injury. gible (FIGURE 1). Although osteochondral lesions may perforate The changes in microenvironment, such as stress and strain, flow velocities, hydraulic pressure and local growth factor concentrations continuously influence the pathologic condition of articular cartilage. Hence, owing to the time from injury to treatment, the dearth of knowledge regarding the natural history of these lesions and the variability in joint health, chondral injury presents a significant challenge for orthopedic research. Specifically, the following procedures and products we will discuss aim to achieve the common goal of shifting ‘cartilage repair’ towards ‘cartilage regeneration’. Until we can reliably restore the biochemical, histological and biomechanical properties of hyaline articular cartilage, the long-term efficacy of current treatment regimens may not be fully recognized.

Cartilage repair procedures There are essentially two approaches to treating articular cartilage injury. The first option, as in debridement, microfracture and mosaicplasty, includes procedures that induce tissue repair Figure 1. Massive full-thickness cartilage defect in the femoral trochlea and/or restoration of the defect using only surgical techniques covering almost the entire articular surface. and instrumentation. The following three sections will focus on

Expert Rev. Med. Devices 4(3), (2007) Page 13

Articular cartilage repair: procedures versus products

these procedures. The second option, surgical implantation of success [26–30]. The multielectrode version, basically diathermy products created in manufacturing laboratories, will be dis- instruments configured for arthroscopic use in a fluid environ- cussed later. Arguments for the application of both approaches ment, comes in various physical shapes with 30 and 90° angles to cartilage repair are reported in the literature. However, carti- for increased surgical simplicity. It is noteworthy that, lage repair procedures offer decreased healthcare costs for although the use of thermal and radiofrequency ablation gives patients; therefore, unless a significant improvement in clinical smoother surfaces compared with motorized debridement via efficacy can be demonstrated, products will always be under an arthroscopic shaver, evidence of thermal necrosis has been greater scrutiny by healthcare providers. reported [26,27,30,31]. High-pressure, fluid-driven burrs have also recently been evidenced to cause less injury to surrounding tis- Arthroscopic lavage & debridement sues and provide smoother debrided surfaces [32–34]. Such Arthroscopic lavage (wash out) of injured synovial joints has devices (e.g., Hydrojet®) show promise in providing minimal long been the primary treatment of chondral lesions [21]. The additional trauma during arthroscopic debridement. procedure removes inflammatory mediators, loose cartilage and In short, while arthroscopic lavage/debridement provides any cartilaginous debris residing in the synovial space causing symptomatic relief to patients by cleaning the joint of loose car- synovitis, effusion or biomechanical pathologies, such as crepi- tilage and debris tissue, as with pharmacologic intervention, it tus. Jackson, a pioneer of arthroscopy, noted that lavage often has not been demonstrated to make any quantitative or qualita- relieved symptoms of joint pain. He reported symptomatic tive advance on defect repair; a necessity to build the defect improvement in 45% of patients 3.5 years after arthroscopy void and impede further articular deterioration. and measurable improvement in 80% of patients [22] – a claim later backed by other research [23]. The debridement, or Microfracture chondroplasty, of loose cartilage was also conducted by arthros- In 1997, Steadman and colleagues documented the use of an copy, with a follow-up study by Jackson, demonstrating 88% awl to create multiple microfractures (MFX) in the sub- initial improvement and 68% prolonged improvement follow- chondral plate of debrided articular cartilage defects in order to ing the procedure [24]. Of course, arthroscopic therapy may accentuate tissue adhesiveness for infiltrating marrow fibrin address patient symptoms (especially in small focal lesions) and clots [35]. An 11-year average follow-up study by Steadman on slow further degeneration by the removal of loose cartilage and full-thickness cartilage defects arthroscopically treated with collagenous debris, but it does not in any way facilitate defect MFX reported significant functional (Lysholm and Tegner repair, nor does it prevent future defect enlargement. This was scores) improvement [36]. However, histologic analysis of illustrated in a classic study by Moseley and colleagues, who microfracture repair has shown, as is the case with all marrow demonstrated that neither arthroscopic lavage nor debridement perforation techniques, that fibrous to fibrocartilage tissue produced improved pain or function compared with placebo dominates the repair site [37]. Furthermore, clinical outcome arthroscopy [25]. has been reported to deteriorate significantly 18 months after The usual debridement equipment used is a standard 30° microfracture surgery [38]. arthroscope, with the occasional use of a 70° arthroscope to The procedure used for MFX is completely arthroscopic and access the posterior femoral condyles. Debridement is carried the equipment utilized is quite basic. The first step is to out by an array of arthroscopic instruments, including arthro- debride the chondral lesion to stable, vertical borders. The scopic punches, straight duck-bill punches, 15° upcutting most common and widely used technique, MFX, was pio- duck-bill punches, basket punches, arthroscopic scissors and neered by Steadman who gave his name to the awl utilized for small rongeurs. However, the most common instrument used is this procedure [35]. The Steadman awl, basically a miniature ice the curette. These come in a variety of sizes and designs, with a pick, is used to produce multiple perforations 3–4 mm apart in ring curette generally being the most useful, especially when the subchondral bone plate. Controlling this distance between biopsy is required. Arthroscopic shavers have also been in use perforations maintains the structural integrity of the sub- for many years and come in various designs, from full-radius chondral bone, therefore minimizing sclerotic bone formation resectors to half-radius resectors, with straight or curved ends. and stabilizing maturation of the infiltrating clot. In addition, These instruments are motorized by either a hand or foot con- it is important not to perforate too deeply as the shape of the trol and have a suction facility attached to facilitate surgical awl means that the deeper the perforation, the wider the perfo- debridement. They can shave in one direction or, more com- ration. Undue care during this procedure can lead to instability monly, can be set on an oscillatory setting in combination with of the subchondral plate and, therefore, macrofracture and a half-radius shaver to improve safety by reducing the chance of subsequent sequelae. snaring adjacent tissue, such as meniscus or synovium. At a cellular level, it has been evidenced that the volume of Recent advances in arthroscopic debridement involve tech- MFX repair and the degree of both type II collagen and aggre- niques and procedures aimed at producing macroscopically can production are less than required to instigate the matura- and histologically smoother debrided surfaces. Ablation of tion of functionally stable cartilage [37,39]. Moreover, cells iso- damaged cartilage with a multielectrode, radiofrequency or lated from MFX repair tissue (compared with biopsied laser arthroscopic wand has been reported to have variable autologous chondrocytes) have been shown not only to lack

www.future-drugs.com Page 14

Willers, Partsalis & Zheng type II collagen expression but also to express osteocalcin, a 1 mm apart. Interestingly, osteochondral autograft transplant- marker of mineralization and bone formation [39]. This lack of ation has also been adapted recently to a retrograde method for type II collagen expression in MFX-stimulated cells was later treating distal and proximal tibial articular defects through a confirmed by Dorotka and colleagues [40]. However, while these tibial tunnel, with satisfactory magnetic resonance imaging MFX-stimulated cells appear to be of poor quality for cartilage (MRI) outcomes reported [49]. One variation on this technique repair, they may be amenable to growth factor manipulation. is the MEGA-osteochondral autograft transfer system (OATS) For example, the addition of bone morphogenetic protein used for large chondral lesions (>2×3cm) [50]. In this tech- (BMP)-7 has recently been shown to improve the repair histol- nique, the autologous posterior femoral condyle is transplanted ogy of cartilage defects significantly compared with MFX to a large weight-bearing defect on the same condyle. However, alone [41]. Although such technology may be promising for given the destructive nature of the procedure, the long-term chondral repair, MFX is still damaging to the subchondral plate impact of donor-site morbidity on knee joint function is ques- and is influenced by the osseous health of the patient. It has tionable. Subsequently, allograft osteochondral cylinders have been demonstrated that there is a reduction in marrow progeni- been investigated as a possible alternate donor source in this new tor cell production with age, in osteoporosis and arthritis [42–44]. treatment regime [51]. This decreased availability of progenitor cells translates into an Commercially available osteochondral autografting systems impaired capacity of MFX to contribute to cartilage repair in include the Acufex® Mosaicplasty™ system, the Arthrex OATS these patient populations and hence limits its demographic system, Depuy Mitek’s COR™ System and the Soft Delivery application. There is also concern regarding the number of System (SDS) by Zimmer. Clinically, while Hangody and col- MFXs warranted per procedure to produce valid repair. This leagues recently reported a 10-year follow-up study of 831 knee number is poorly regulated and varies between surgeons and mosaicplasty patients, demonstrating good-to-excellent results with the size of the defect. While more MFXs allow more pro- in 92% of patients treated with femoral condylar implantations genitor cells to infiltrate the defect and facilitate chondro- and 79% in patellar and/or trochlear mosaicplasty patients, the genesis, they may also destabilize the subchondral plate, com- technique has not been favored in the literature owing to its promising immediate restoration of osteochondral architecture surgical difficulty and subsequent variability in outcome [52,53]. and biomechanical transduction through the joint. A novel technique was described recently utilizing Since little information has been published on the biological computer-navigated, arthroscopic osteochondral grafting in a mechanism of MFX, understanding the variable outcomes of cadaveric model of chondral defect in the talus [54]. In this pro- this technique is difficult. Indeed, many factors govern the bio- cedure, a computer-navigated guide-wire was passed retrograde logical success of MFX. For example, the unpredictable pheno- into the talus, and the chondral surface was prepared using type of the infiltrating reparatory cell population. The creation cannulated reamers and arthroscopic mechanical shavers. of these MFXs not only introduces MSCs of variable pheno- Osteochondral grafts were obtained from the lateral trochlea typic profiles from the marrow but may also contaminate defect and ipsilateral femur, then inserted in retrograde fashion, repair with fibroblasts and osseous debris from the disrupted chondral surface first, through the body of the talus to recon- region. The organization of these infiltrating cells is also impor- struct the articular surface. While only a cadaveric model, this tant to the function of articular cartilage. Articular cartilage type of approach to mosaicplasty does stimulate thought on matures over years of complex development via endochondral further applications of this method although, as with ossification and intramembranous bone formation, in conjunc- conventional mosaicplasty, the imperative behind further tion with numerous physical and chemical factors. Therefore, development must be precise plug fixation. we must replicate many of these factors to reproduce healthy Indeed, surface discontinuity (due to unstable osseous foun- and functionally stable cartilage that will last. dations and/or structural donor–recipient differences), unmatched thickness of donor and recipient cartilage, differen- Osteochondral autografting (mosaicplasty) tial structural orientation of the cartilage matrix and weak The most obvious method for filling a chondral defect void fibrocartilage grouting are some examples of complications would be a biocompatible graft of the same tissue characteristics. resulting from this method of cartilage repair. Most impor- Developed by Matsusue, Hangody and Bobic, osteochondral tantly, care must be taken to match donor graft depth to the autografting has been thought by many orthopedic surgeons in reception site, to match the thickness of donor and recipient the past to be the gold standard in cartilage repair [45–47]. The cartilage and to ensure that donor plugs are removed and trans- first step in osteochondral autografting is the debridement of the planted perpendicular to the surface to maximize defect to produce a well-circumscribed site for graft transplant- donor–recipient structural homogeneity. Wu and colleagues ation and a subchondral surface capable of graft adhesion. The recently reported that minor surgical displacements of the grafts are extracted from either the supracondylar ridge or the osteochondral plug induce abnormal tension in the opposing superior intercondylar groove using bone punches (usually 4.5 healthy articular cartilage surface, which may have patho- or 6.5 mm) the same diameter as the drilled recipient holes, and mechanical or degenerative sequelae [55]. The matching of graft inserted into the drilled defect using a graduated harvesting length and recipient drill hole length has also been highlighted tamp [48]. Grafted cores are generally inserted approximately as an important determinant in successful graft insertion. For

Expert Rev. Med. Devices 4(3), (2007) Page 15

Articular cartilage repair: procedures versus products example, minimal countersinking (1 mm) of plugs in sheep has been reported to facilitate chondrocyte hyperplasia, tidemark advancement and osseous integration, while countersinking to 2 mm produced cartilage necrosis and fibrous overgrowth [56]. Biomechanically, it has also been illustrated recently that increased plug diameter infers greater primary stability following transplantation and that there is no difference in stability between implanting plugs in a circular or linear manner [48]. Additionally, the elevated force needed to insert plugs longer than their drill hole has been linked to disruption of the donor cartilage and increased chondrocyte death [57,58]. Chondrocyte death has been reported to occur in the plug circumference due to the trauma of harvesting; a problem exacerbated by the use of a power trephine [59,60]. Chondrocyte cell death through wounding has been suggested to be caused by necrosis and apoptosis, with variation reported within samples [60,61]. Indeed, Tew and colleagues noted that some cells die instantly following wounding, while immediately neighboring cells may proliferate and apoptosis may increase over time [61]. In short, until a stable nutritional supply to the plug chondrocytes is re- established, cells may autophagocytose, abutting cells in order to acquire nutrients to maintain cell biogenesis. Regardless of the cause, a lack of matrix-producing chondrocytes at the cartilage–cartilage transplant junction or within the graft makes plug integration and survival, and hence, functionally adequate repair, highly improbable (FIGURE 2). Owing to the observed variability in histological, bio- Figure 2. Histology of poor mosaicplasty repair integration at 1-year mechanical and clinical outcomes of osteochondral auto- follow-up. A clear space is visible between the transplanted osteochondral grafting, this procedure still requires substantial follow-up in plug (right) and the native articular cartilage (left). The transplanted plug also order to evaluate the integration of the grafted tissue and the appears to have deteriorated, with fibrillated edges and a predominantly biomechanical competence of the graft site. While the out- fibrocartilaginous matrix morphology. comes of this method appear promising, mainly for small defects, the aforementioned technical difficulties and sub- cell–scaffold products for the treatment of cartilage defects, sequent complications pose significant questions regarding the given the elevated cost to patients, the comparative efficacy of reproducibility and longevity of a successful clinical outcome. these has yet to be established.

Cartilage repair products Carbon fiber implantation The second cartilage repair option, as in matrix-induced auto- One of the first attempts at repairing articular cartilage lesions logous chondrocyte implantation (ACI), is the surgical implan- of the knee entailed the implantation of carbon fiber rods (CFI) tation of prefabricated products created in approved manufac- and pads into the affected area. The idea of chondral resurfac- turing laboratories for the treatment of cartilage defects. There ing was first introduced in 1978 using an experimental animal have been many developments in products for cartilage regen- model, with subsequent results showing a predominance of eration over the past few decades. The first product for carti- fibrous tissue filling with good surface continuity [62,63]. Some lage repair, carbon fiber implantation, was introduced in the studies have even reported the production of hyaline cartilage, 1970s. However, most current research has surrounded the albeit with the accompanied implantation of cultured cells or retrieval and implantation of autologous chondrocytes, with periosteum [64,65]. Brittberg and colleagues reported 83% good many variations in surgical technique and cell-delivery systems or excellent results in 37 patients, noting pain relief as an described. TABLE 1 shows the main ACI products currently avail- important result [66]. Carbon fiber has also been used to fill the able, countries using them, longest follow-up period and peer- donor sites created during mosaicplasty, but was noted to stim- reviewed publications (PubMed search methods: keywords – ulate little repair tissue formation by 30 weeks [67]. However, ‘autologous and chondrocyte’; limits – humans only, original studies supporting the use of CFI have faded over the last articles only). This in vitro fabrication and subsequent implan- decade, with reports arising of postoperative pain and discom- tation of cartilage tissue using a combination of cells and bio- fort, foreign body giant cell reaction, fiber debris deposition, scaffolding materials has been termed ‘tissue engineering’. prevention of the restoration of subchondral and osteochondral While the field of tissue engineering has produced many architecture and low patient satisfaction compromising support

www.future-drugs.com Page 16

Willers, Partsalis & Zheng

Table 1. Autologous chondrocyte implantation products: countries, companies, number of peer-reviewed references and maximum follow-up. Product Country Company Number of peer- Maximum reported reviewed publications follow-up (years) Periosteal ACI USA Genzyme 53 11 UK BioTissue Germany Cellontech Australia Tigenix Switzerland Educell Belgium Codon Sweden Tetec Austria Norway Korea Matrix-induced ACI (MACI®) USA Genzyme/Verigen 13 5 UK Arthrex Australia Denmark Italy Spain Collagen ACI UK Cellgenix 8 7 Australia Geistlich Germany Genzyme/Verigen Ormed Hyaluronan ACI Italy Fidia 7 5 Collagen gel ACI Germany Ars Arthro 1 4.7

ACI: Autologous chondrocyte implantation. for the product in cartilage repair [68–70]. However, a recent basically consists of the following surgical and laboratory com- 5-year follow-up study of CFIs has reported good tolerance and ponents: arthroscopic chondrocyte harvest; in vitro chondrocyte osseous bonding of residual carbon fibers [71]. The biggest issue cultivation; open surgery defect debridement, excision, sizing with CFIs is that the fibers, whether rods or pads, do not and suture of tibial periosteal flap; and injection of cultured appear to allow tissue infiltration and are not significantly chondrocytes. In 1984, Peterson was the first to demonstrate absorbed by the body in order to allow local repair mechanisms that periosteal ACI (ACI-P) could be used to treat chondral to re-establish the native tissue architecture. Owing to these injury, presenting the successful transplantation of autologous factors, the use of carbon fiber in cartilage repair has declined chondrocytes into chondral patellar defects created in rabbits over the past decade, with the focus shifting towards biological [72]. The project was later described in the 1989 publication by tissue-engineered cartilage resurfacing. Grande and colleagues [73]. In 1994, the same Swedish group documented the successful treatment of deep cartilage defects in Autologous chondrocyte implantation 23 patients using ACI-P [74], with 87% good and excellent Although conventional ACI may be considered a cartilage repair results in femoral condylar repair and 73% demonstration of procedure in some ways, the commercial requirement for hyaline-like cartilage upon microscopy. Subsequently, a chondrocyte cultivation between cartilage harvest and implant- 2–9-year follow-up of the first 100 ACI-P patients demon- ation inevitably classifies ACI as a product. Conventional ACI strated 96% good and excellent outcomes in focal femoral

Expert Rev. Med. Devices 4(3), (2007) Page 17

Articular cartilage repair: procedures versus products

condyle lesions, 89% in osteochondritis dissecans patients and (ICRS) assessment [89]. Following on from this finding, 75% in ACI reconstruction-related defects [76]. It should be Haddo and colleagues also demonstrated clinical improve- noted, however, that the histological characterization of ACI-P ment at 2 years post-ACI-C [87]. Biologically, further basic repair at 2 years is variable in remodeling and matrix synthesis research has demonstrated that ACI-C implanted chondro- status [75]. It should be noted however, that ACI-P in the patel- cytes express type II collagen at the same time period [85]. lofemoral joint has been shown to be less effective than treated However, while ACI-C is at the very least a technical improve- condylar and trochlea lesions [74,77]. Furthermore, the condition ment on ACI-P, the need to suture the membrane to the defect of patella tracking is thought to determine the clinical success of circumference is not desirable (FIGURE 3). ACI-P in the patella [78]. Additionally, the regenerative capabilities, larger surgical Matrix-induced ACI ® incision, peripheral graft hypertrophy [79–81], graft delamina- As in ACI-C, matrix-induced ACI (MACI ) also uses collagen tion [80–83] and possible calcification problems [80,84] associated membrane as an integral component. However, in MACI, with periosteum should not be overlooked when considering chondrocytes are seeded directly onto a type I/III collagen mem- periosteal ACI as a treatment option. Indeed, a recent report brane instead of being added to the defect as a cell suspension. on the adverse events of ACI-P by the US FDA cited 3.8% of Then, in conjunction with fibrin sealant, the seeded membrane patients with an adverse event in the total manufacturer distri- is adhered to the base of the defect (FIGURE 4) to instigate chon- bution (7500 Carticel lots) from 1995 to 2002, with dral repair. MACI uses a type I/III collagen membrane seeded 497 adverse events (294 patients from 1996 to 2003), and with cultured autologous chondrocytes to regenerate articular more than one adverse event was noted in 46% of these cartilage defects. Bypassing the need for microsuture, fixation of patients [79]. Adverse events comprised 25% graft failure, 22% the chondrocyte-seeded membrane is achieved using fibrin seal- graft delamination and 18% repair tissue hypertrophy, with ant injection onto the subchondral bone bed of the defect. The 96% of events involving the femoral condyles. Similarly, Minas membrane is then shaped to match the defect geometry and has reported that complications linked to the use of periosteum gently pressed into the sealant to assure graft adhesion. The in ACI may occur in 20–25% of patients, while others have fibrin sealant acts as an adhesive substrate and possibly promotes reported up to 87% lifting or delamination of the periosteal the migration and biosynthesis of chondrocytes from the mem- ® graft [80,82]. These issues lead to reoperation or the need for brane [90]. Brittberg reported that fibrin sealant (Tisseel ) inhib- arthroscopic shaving of the graft’s edges to maintain cartilage its chondrocyte migration into chondral defects; however, surface continuity. Although it is unclear whether hypertrophy numerous subsequent studies, including the authors’ own has a significant impact on the clinical outcome of ACI-P, the research, contradict this lone result [90–93]. Furthermore, it is esti- need for revision surgery is certainly undesirable. Furthermore, mated that more than 4000 MACI operations using fibrin seal- the need to microsuture the periosteal patch to the defect ant as a graft adhesive have been conducted worldwide with no boundary is also a technical concern, with the damage caused obvious clinical complications reported to date. In the bilayer at the suture site and the limitation to patch integration always collagen membrane, however, the cambium side provides an a compromising component of ACI-P repair. Hence, while ACI-P has evidenced good clinical effectiveness, the various complications associated with its use of periosteum in cartilage repair have troubled surgeons and encour- aged researchers to investigate alternate bioscaffolds and delivery systems for ACI.

Collagen-covered ACI If one was to conduct a thorough review of the literature, collagen-based bioscaffolds appear to have gained high popularity for cell-based cartilage repair. There are currently a number of publications advocating the use of collagen membrane as a component of ACI [85–88]. By far the greatest advantage to using collagen membrane is the obviation of reported periosteal-associated complications. As the type I/III collagen membrane is acellular and, hence, seemingly inert within the joint, it is not predisposed to the same hypertrophic overgrowth commonly observed in collagen-covered ACI (ACI-C). Bartlett and colleagues have illustrated significant clinical Figure 3. Medial femoral condyle defect treated by collagen-covered autologous chondrocyte implantation. Microsutures can be seen radiating improvement (p = 0.01) 1 year after ACI-C with the Cincin- from the periphery of the defect, with fibrin sealant being added to seal the nati Knee Score, and 79% good-to-excellent repair appearance previously injected chondrocyte solution. under arthroscopic International Cartilage Repair Society

www.future-drugs.com Page 18

Willers, Partsalis & Zheng

colleagues noted a 50% increase in chondrocyte proliferation after HA addition in comparison with controls [100]. Further- more, the implantation of autologous chondrocytes aboard a HA derivative bioscaffold (Hyaff®) produced significantly better regenerative tissue in a rabbit compared with bioscaffold alone and untreated controls [99]. Such polymer-delivery systems are also favorable owing to the capacity for arthroscopic implantation. One group has reported improved postoperative results and no complications after arthroscopic implantation of the tissue-engineered hyaluronic graft [101]. Fidia Advanced Biopolymers, an international tissue engineering company, has also shown excellent clinical results using its HA scaffold (Hyalograft® C) to implant autologous chondrocytes, with a reported 94% improvement of quality of life (QoL), 87% improvement in functional outcomes and the majority of samples presenting Figure 4. Large cartilage defect in the femoral head treated by hyaline-like repair [103]. More recently, Hyalograft C has been matrix-induced autologous chondrocyte implantation (AMCI®). As shown to produce no adverse events and clinical improvement illustrated, the MACI product allows effective shaping of the graft to the at 3 years’ follow-up, with a particular indication for young defect site, without the need for microsuture. patients with single lesions noted [104]. It has also recently been ideal environment for chondrocyte attachment and matrix syn- illustrated that Hyalograft C has the capacity to regenerate hya- thesis (FIGURE 5), the smooth side mimics the native joint surface line cartilage within osteoarthritic knees, a phenomenon also for unrestricted articulation. Indeed, early complete adherence of noted by the authors [105]. At a molecular level, HA scaffold the MACI graft has been shown to occur in over 80% of (Hyaff®-11) has been shown both to increase hyaline-specific patients, and the implanted cells have been shown to be of high matrix synthesis and attenuate the onset of OA through the viability and express markers for differentiated hyaline chondro- regulation of anabolic and catabolic factors [106]. Specifically, cytes [94,95]. These findings commonly translate to hyaline-like Hyaff-11 upregulated type II collagen, aggrecan and SOX-9 cartilage repair with good tissue integration (FIGURE 6). Interest- levels, downregulated matrix metalloproteinase (MMP)-1 and ingly, some researchers have trialled a sandwich variation of the nitrite levels, and decreased chondrocyte apoptosis with time. MACI technique involving the use of two chondrocyte-seeded These data support the use of HA scaffolds in chondral injury, membranes, but this has certainly not gained any popularity in while also suggesting a benefit in early OA patients. the literature [96]. Cherubino was one of the first to support the MACI technique, Collagen gel ACI publishing improved clinical and functional outcomes, no com- Similar in nature to HA, type I collagen gel is a 3D bioscaffold- plications and MRI-visualized hyaline-like cartilage [86]. Addition- ing material trialed as a carrier for chondrocyte implantation ally, a recent 5-year follow-up of MACI patients has reported sig- into chondral lesions [100,107,108]. As with other 3D gel nificant clinical improvement, advocating the product as a suitable but cost-intensive alternative in cartilage repair [97]. Currently, little complication following MACI has been reported in the literature after more than 8 years of clinical experience. In fact, at 5-year follow-up, no instance of hypertrophy or calcification of the graft was reported [97]. However, as with all ACI variant techniques, the long-term efficacy of cartilage repair may not be realized for some time and longer follow-up studies must be conducted to validate this product.

Hyaluronic acid ACI Hyaluronic acid (HA) is a linear polysaccharide that is a major component of hyaline articular cartilage extracellular matrix. The therapeutic potential and ability of HA to differentiate chondrocytes has been supported by various studies [98–103]. In vitro research on the effects of HA on chondrocyte behavior Figure 5. Autologous human chondrocytes aboard a type I/III collagen has demonstrated that it induces the phenotypic expression of membrane after 3 days of culture. Cells readopt their differentiated collagen II and proteoglycan molecules, as well as stimulating spherical morphology after monolayer culture and carpet the scaffold congruously with the underlying collagen matrix. cellular proliferation [98–100]. To this end, Kawasaki and

Expert Rev. Med. Devices 4(3), (2007) Page 19

Articular cartilage repair: procedures versus products constructs such as HA, type I collagen gel has been evidenced to maintain the differentiated chondrocyte phenotype for implantation into chondral defects [107,108]. Interestingly, HA has been reported to aid the chondrogenic effects of collagen gel by enhancing chondrocyte proliferation and chondroitin sulfate production [100]. Although it should be noted that significant (55%) diametric contraction of floating chondrocyte gel constructs has been reported, it is unclear whether this phenomenon may compromise the integration of such constructs follow- Native cartilage ing implantation [107]. It should also be noted that conventional collagen gels can be problematic, owing to immunogenicity and Hyaline-like repair safety concerns. Atelocollagen® gel is another commercially available collagen gel. In the production of the gel, the telopeptide antigenic components on the peptide chain of type I collagen are removed [109]. The source of atelocollagen is generally bovine Figure 6. Hyaline-like cartilage repair, with seamless integration skin, and the gel has been used clinically for the treatment of between the repair and native cartilage (center). Image acquired from dermatological disease, as well as being utilized as a carrier for a matrix-induced autologous chondrocyte implantation patient at BMP for osseous repair [110,111]. The effectiveness of atelo- 12 months postoperatively. collagen in cartilage has been suggested by a number of groups, both in preclinical research and the clinical setting [112–116]. One requirement for two procedures: cell harvest and implantation. of the more interesting preclinical studies using rabbits evi- The most feasible way to bypass this issue would be to create a denced that chondrocytes cultured in atelocollagen gel facili- cell bank of allogenic chondrocytes for multipatient implant- tated stable repair integration and tissue thickness, compared ation. This would provide a plentiful source of chondrocytes and with those implanted with monolayer cultured chondrocytes would help to satisfy the increasing demand for ACI. Of course, that demonstrated tissue degeneration with time [115]. It has also the major concerns surrounding such an idea are those of been illustrated that the impact of these chondrocyte–atelo- immunogenic compatibility and disease transmission. Few pre- collagen constructs is influenced by cell density, with increasing clinical studies have looked at allogenic chondrocyte implant- density producing more cartilaginous tissue but less cell prolif- ation but those that have generally document no complications eration [114]. Such information must be considered when and improved cartilage repair than controls [118–121]. One of the developing these technologies for clinical application. Clinically, biggest concerns regarding the use of allogenic chondrocytes is autologous chondrocytes cultured in atelocollagen gel have been the ability for them to integrate with the surrounding cartilage in reported to improve functional outcome significantly, with a cartilage lesion. To this end, Kim and colleagues have evi- 93% good-to-excellent arthroscopic grading and cartilage stiff- denced that allogenic chondrocytes attach to explant cartilage, ness similar to that of the surrounding cartilage at 2 years [113]. proliferate and produce matrix positive for collagen II and glyco- Additionally, it has been evidenced recently in chondral and saminoglycan at 4 weeks [120]. Furthermore, Wakitani and col- osteochondral femoral defects that treatment by type I collagen leagues, although using broad histological criteria, reported 80% gel seeded with autologous chondrocytes (CaReS®) significantly complete cartilage repair over 24 months in the rabbit knee fol- improves International Knee Documentation Committee lowing implantation of allogenic chondrocytes within collagen (IKDC) and functional scoring at 2 years, with 84% of patients gel [119]. Also, allogenic chondrocytes have been documented to rating their outcome as good to excellent [117]. Although, as produce hyaline cartilage in horses, with significantly increased with other 3D chondrocyte implantation techniques, atelo- glycosaminoglycan (GAG) and type II collagen content at collagen appears to maintain phenotypically stable chondrocyte 8weeks [122]. But while allogenic chondrocytes appear to be populations and allow arthroscopic treatment of chondral effective in animals, no clinical reports of allogenic chondrocyte injury, these constructs require further validation before implantation exist to date. widespread application is justified. Stem cell-based cartilage repair Allogenic chondrocyte implantation Other than chondrocytes, few cell lines have been studied for Although mainly associated with medicolegal and ethical issues, their chondral regenerative capabilities. Most research other than allogenic chondrocyte implantation has not yet been widely ACI has focused on MSCs as an alternate source for cartilage accepted for clinical use, but this treatment modality is certainly repair. The potential of MSCs for cartilage repair was first the most patient, surgeon and economically friendly at present. described by Wakitani in 1994 [123]. In a rabbit model, Wakitani While the aforementioned ACI methods using biodegradable isolated MSCs from marrow or periosteum to treat chondral knee membranes and gels produce histological and clinical improve- defects. The implanted MSCs distributed uniformly throughout ment, albeit variable, their overall efficacy is undermined by the the defects to produce subchondral bone and cartilage similar to

www.future-drugs.com Page 20

Willers, Partsalis & Zheng native, with 24-week repair tissue stiffer and less compliant than hyaline articular cartilage. However, while preclinical stem cell untreated defects, but less stiff and more compliant than healthy studies have suggested a therapeutic potential for this technol- articular cartilage. This work was later supported by Grande, who ogy, the ethical approval of these cells in human trials is still a similarly advocated MSCs as an effective therapeutic alternative distant objective. for the treatment of cartilage defects [124]. Subsequently, various studies, both in vitro and in vivo, have Comparison of efficacy between cartilage repair procedures advocated the use of MSCs for chondral repair [124–128]. These & products cells have been isolated from tissues, including bone marrow, In general, the reported variability in outcomes within the lit- periosteum, muscle, adipose tissue, cartilage and synovial mem- erature and the lack of long-term randomized controlled stud- brane for differentiation assays in vitro or in vivo cartilage ies on the many cartilage repair procedures and products make repair [127–133]. Interestingly, one recent paper has suggested that any high-quality analysis of their comparative efficacy difficult MSCs may be able to be sourced and differentiated from at best. There are, however, some comparative studies worthy human placenta, thereby expanding the potential of these cells of note. for cartilage repair [134]. However, a review of the literature In terms of cartilage repair procedures, one recent study has reveals qualitative differences in the repair tissue outcomes compared all three of the aforementioned cartilage repair pro- across varying stem cell sources. Nathan and colleagues reported cedures in the ankle with a mean follow-up of 53 months [139]. superior histological grading in adipose-derived stem cell repair Interestingly, the results demonstrated no difference between compared with periosteum-derived stem cell repair in the same chondroplasty, MFX and osteochondral autografting with regard study [131]. The study also noted that stem cells were more read- to patient function; however, pain 24 h postoperatively was sig- ily obtained from adipose tissue than bone marrow or peri- nificantly lower in chondroplasty and MFX patients. In contrast osteum. Given the relative abundance of adipose tissue in the to this study, Gudas and colleagues found osteochondral body and the ease of acquisition, it may make sense to target autografting superior to MFX in young athletes, with only 52% this cell source in future orthopedic stem cell research. returning to sports at the preinjury level after MFX, compared Nevertheless, while there are various sources to isolate MSCs, with 93% following osteochondral autografting [140]. the obvious primary challenge for using MSCs in cartilage Perhaps the most contrasting of the cartilage procedures and repair is to differentiate them into phenotypically stable products are debridement and MFX versus ACI, respectively. chondrocytes. If the biological function of these cells is not reli- Fu and colleagues compared debridement and ACI at 3 years able, we cannot reliably expect qualitative tissue regenerative follow-up, documenting that 81% of patients improved their and clinical outcomes, nor can we guarantee the safety of such overall condition after ACI compared with 60% after debride- products. Indeed, Zheng and colleagues have commented on ment, greater comparative improvements in pain and swelling the need to regulate the stability of human cell-based (MSC) after ACI, and the same failure rate for the two techniques [141]. products (Class 4 products under Australian regulations) in In terms of MFX, a randomized trial in 2004 by Knutsen and order to provide a quality standard [135]. This has mainly been colleagues compared 80 patients having either MFX or ACI tackled on two fronts. First, the effects of various growth factors [142]. The results demonstrated that at 2 years both techniques on driving MSCs into the pathways of chondral differentiation had improved and comparable clinical outcomes (ICRS, have been noted over recent years [127,132,136]. For example, Lysholm, SF-36 and Tegner forms), with only a significant muscle-derived stem cells (MDSCs) have been suggested as a improvement in the MFX SF-36 physical component score possible source for stem cell-based cartilage repair. Interestingly, noted. The study concluded that the repair histology achieved MDSC chondrogenesis has recently been shown to be amena- was not significantly different between the groups and could ble to retroviral transduction with BMP-4, with improved find no correlation between the histological grade of the repair chondrogenic phenotype and cartilage defect repair noted in and the associated clinical outcome. Interestingly, Knutsen also 2 comparison with those transduced to express BMP-4 [127]. Also, noted MFX patients with smaller lesions (<4 cm ) had signifi- it is well known that certain 3D scaffolds induce chondro- cantly better clinical results (p < 0.003) than those with larger genesis; hence MSCs have been incorporated into these materi- lesions, but did not find any association between lesion size and als for differentiation. A recent study demonstrated that MSCs clinical outcome in ACI patients (p > 0.89). This suggests that grown within a 3D scaffold (Hyaff-11) exhibit reduced expres- ACI techniques may be indicated more for larger symptomatic sion of catabolic factors, while providing a good environment lesions of the knee than MFX. By contrast, the new generation for chondrogenic differentiation [125,137]. Furthermore, MSCs of ACI, MACI, has been reported by Basad to show a signifi- implanted within polyglycolic acid (PGA) into rabbit femoral cantly (p = 0.049) better functional outcome (Lysholm–Gil- defects have illustrated stable cartilage repair up to 42 weeks lquist score) compared with MFX at 24 months [143]. However, after surgery [138]. the Tegner–Lysholm and ICRS scores at 24 months were not A number of groups have proven that MSCs can be isolated significant (p = 0.064 and p = 0.32, respectively). Furthermore, and cultured without compromising their differentiation another study investigating the impact of cell type in MFX by capacity; however, the true challenge is to mimic the complex introducing cultured autologous chondrocytes, has suggested developmental cues that enable the maturation of functional an inferior contribution of MFX-stimulated marrow cells

Expert Rev. Med. Devices 4(3), (2007) Page 21

Articular cartilage repair: procedures versus products compared with autologous chondrocytes in cartilage repair; effective treatment option for cartilage lesions of the documenting superior defect infill and tissue repair quality in knee [147,148]. While randomized controlled trials with a longer cases of autologous chondrocytes combined with MFX [40]. follow-up period may shed more light on the most efficacious Hence, given the unpredictable phenotype of the infiltrating method for cartilage repair in the future, MACI appears to offer marrow cells, this is certainly a valid comparative rationale why comparable clinical outcome and reduced complication rates to ACI is thought by many to be a more reliable treatment choice. the other techniques available. ACI has been advocated for cartilage repair over other conventional methods. Notably, a prospective, randomized, com- Impact of concomitant joint surgery parative study illustrated the advantage of ACI-C over mosaic- As synovial joints, such as the knee, are maintained by complex plasty, demonstrating 89% good-to-excellent functional interactions of various tissues, biochemical factors and biome- outcome (Cincinnati and Stanmore scores) after ACI-P, com- chanical forces, articular cartilage is at risk of injury through a pared with 69% after mosaicplasty [52]. Arthroscopic analysis at number of mechanisms. In particular, abnormalities in the load- 1 year also showed good or excellent repairs in 82% of ACI ing axis of the knee are a strong determinant of the fate of articu- patients compared with 34% in mosaicplasty patients. The lar cartilage injury. Some common procedures used to correct study further commented that mosaicplasty repair deteriorates these biomechanical issues include anterior cruciate ligament with time and is contraindicated for patella lesions due to dif- (ACL) reconstruction, patellar realignment, meniscectomy and ferences in donor and recipient cartilage thickness. A later study high tibial osteotomy. Indeed, Browne and colleagues, in report- contradicted these findings, reporting complete recovery ing 5 years’ experience with ACI, noted 72% concomitant surgi- (Lysholm 90–100) in 88% of mosaicplasty patients compared cal procedures at arthroscopy and 21% at implantation [149]. with 68% in ACI; however, this was ACI-P and the difference However, the contribution of these procedures to the aforemen- was not significant [144]. tioned cartilage repair methods has undergone little investigation As previously discussed, ACI-P has many issues associated with within the literature. the use of periosteum. This was recently illustrated in a compara- ACL deficiency, particularly common in sports medicine, is tive study of the two techniques, which reported that 36% of known to correlate with cartilage injury in the affected joint. A ACI-P cases required revision shaving of hypertrophic tissue, recent paper has reported on the combined treatment of ACL whereas no ACI-C cases required revision [88]. Although having injury and associated cartilage injury [150]. The study docu- the technical advantage of using collagen membrane to reduce mented improved clinical outcomes in patients simultaneously the risk of complication, ACI-C was also shown to produce sim- treated by ACL reconstruction and MACI compared with ilar clinical outcomes to ACI-P. However, as with the ACI-P MACI-treated patients who had previously undergone ACL periosteal patch, ACI-C still carries the technical disadvantage of reconstruction. Similarly, improved results in cases of combined the need for the collagen membrane to be microsutured to the ACL reconstruction and osteochondral autografting have been defect border. reported [47]. The MACI technique is superior to both ACI-P and ACI-C as As patella ACI outcomes are typically variable in comparison it eliminates the high periosteal complication rate and improves with femoral patients due to patellofemoral dysfunction, many surgical simplicity by obviating the need for microsuture. Haddo surgeons choose to unload the patellofemoral joint during carti- and colleagues evidenced the advantage of collagen membrane lage resurfacing surgery. Complementary treatment of ACI and over periosteum, documenting good arthroscopic results and no extensor mechanism realignment has been advocated for hypertrophy at 1 year compared with periosteum [87]. However, improving ACI outcomes by a number of studies [76–78,151]. while the MACI technique has improved the complication rate Minas evidenced poor outcomes in type III and IV patella injury and surgical simplicity of this treatment product, it has not yet in osteotomy alone patients, compared with those with accom- been shown to produce a significant improvement in clinical out- panying ACI [77]. More recently, Henderson has reported signifi- come in comparison to either ACI-P or ACI-C. To this end, Bar- cant clinical improvement at 2 years’ follow-up in patients hav- tlett and colleagues have reported comparable clinical, arthro- ing concomitant ACI and extensor realignment compared with scopic and histological outcomes between MACI and ACI-C, those with untreated patellofemoral tracking having ACI while Fu has also documented no significant difference in the alone [78]. These studies suggest that, with or without patella rate of improvement of the two techniques [89,145]. tracking abnormalities, unloading the patellofemoral joint trans- There are currently few comparative data on hyaluronan lates better results following ACI and may help to maintain the ACI and collagen gel ACI against other ACI methods, durability of the repair tissue. although one study has reported no difference in IKDC scor- Cartilage injury in the knee also presents subsequent to ing in collagen gel ACI (CaReS) versus conventional ACI and meniscal tear and, vice versa, a severely damaged articular sur- shorter operation times with CaReS (69 vs 107 min) [146]. face can predispose meniscal tear [152]. The most common In summary, two systematic reviews of the literature have treatments of meniscal tear are meniscal repair and menis- concluded that no evidence currently exists to definitively advo- cectomy. Few studies have been conducted into the benefit of cate ACI (or any other cartilage repair product) or conventional these techniques when accompanying cartilage repair but, procedures, such as mosaicplasty or microfracture, as a more given the important shock absorption relationship of the

www.future-drugs.com Page 22

Willers, Partsalis & Zheng meniscus to the opposing articular surface, it is essential that Five-year view any meniscal tear be addressed before cartilage repair is induced Although chondrocyte-based therapeutic products show the by the aforementioned techniques. Likewise, if varus/valgus most potential in cartilage repair at present, there are several malalignment of the tibiofemoral joint compromises condylar avenues of research that will need to be expanded upon in order or tibial plateau cartilage repair, a high tibial osteotomy should to realize the full therapeutic capacity of these technologies. The be performed to unload the affected compartment and protect main research directions worthy of discussion are the precondi- the repair site against excessive forces during weight bearing. tioning of implanted chondrocytes, the development of cell- Additionally, associated ligamentous deficiency should be seeded scaffolds for arthroscopic delivery and the study of allo- treated to restore stability and guard the articular cartilage genic chondrocytes as an alternate cell source. Also requiring against excessive shear forces. investigation is the long-term analysis of the cost–effectiveness of the current array of cartilage repair products and procedures. Expert commentary One of the greatest inherent limitations of treating articular In the short term, most cartilage repair patients will present cartilage injury by cell-construct techniques is the possible varia- with clinical improvement over time. However, while this may bility of cell phenotype inferred by the lack of preimplantation infer symptomatic relief and functional restoration to the conditioning of the introduced cells. We cannot expect to simply patient, long-term follow-up study is essential to reliably add cells to a chondral defect and have them organize hyaline establish the differential efficacy of these procedures and prod- matrix in a controlled manner, nor should we rely on the pheno- ucts to better inform surgeons of the most appropriate treat- typic stability of these cells. Although the utilization of 3D bio- ment option. In saying this, the reliability of current pub- scaffolds has afforded us some control over the distribution and lished studies has also been questioned recently [153]. Jakobsen phenotype of implanted chondrocyte populations, conventional and colleagues reported on the quality of 61 cartilage repair ACI methodologies make little use of known chondrogenic fac- studies, finding several methodological deficiencies [153]. Spe- tors to improve chondrocyte functioning for implantation. Work cifically, a mean methodology score of 43.5 out of 100 was by Luyten and colleagues may help to clarify this issue. Their stated and several shortcomings, such as the type of study, study of predictive molecular markers for stable cartilage forma- description of rehabilitation protocol, outcome criteria, out- tion, together with the ‘ChondroCelect™’ ACI method, which come assessment and subject selection process, were noted. uses chondrocytes grown in specialized media to allow predictable Furthermore, a total of 27 clinical outcome measurement chondrocytic restoration of hyaline cartilage matrix (commercial- scales were used in these studies to assess outcome; a factor ized by Belgian company Tigenix), have been consolidated into a further rationalizing the caution with which many cartilage clinical multicenter randomized control trial. The trial looked at repair studies must be approached. At present, given the cur- ChondroCelect compared with MFX in over 100 patients [201]. rent climate of cartilage repair research, the most efficacious With 1-year biopsy histology and MRI, plus 2–5-year clinical method of cartilage repair appears to be MACI, given the follow-up, the study may shed light on the clinical effectiveness of comparable patient outcomes observed and simpler surgical preimplantation cell conditioning. Indeed, Tigenix have recently technique and reduced complication rate compared with other reported positive Phase III results (at 12 and 18 months), with procedures and products. However, theoretically, the most improvement in repair histology (p < 0.05) and clinical outcome promising method for repairing articular cartilage defects may compared with MFX noted [201]. The next best answer to this lie in the development of arthroscopic allogenic (neonatal) issue is to precondition the chondrocytes in vitro into a tissue that chondrocyte implantation, a technique that would bypass resembles native hyaline matrix as closely as possible. To this end, most biological and economical issues currently limiting the the most important factor missing from in vitro cultivation is the application of ACI. Although the clinical safety and efficacy of biomechanical stimulation of the chondrocytes. The relationship this technology requires further investigation (human leuko- between biomechanical stimuli and the phenotypic conditioning cyte antigen typing may be necessary preimplantation), pre- of articular chondrocytes is a well-accepted relationship. In partic- liminary research has reported that allogenic chondrocytes do ular, bioreactors have been implemented in many studies to assess not elicit an inflammatory response, possibly due to immuno- the impact of biomechanical loading on chondrocyte protection by their surrounding dense extracellular behavior [157–159]. Marlovits and colleagues demonstrated that matrix [154–156]. Moreover, if cartilage repair products are to dedifferentiated monolayer chondrocytes could be redifferenti- become widely used, it must be recognized that the three ated to express collagen II and other hyaline-specific matrix pro- stages (FIGURE 7) of cell-based tissue engineering (product man- teins upon transfer to a bioreactor environment [157]. In general, ufacture, surgeon implantation and patient rehabilitation) static and fixed dynamic compression tend to decrease matrix must all be performed at a high standard in order to realize the synthesis over time, whereas shearing compression and oscillatory most efficacious outcomes. Regardless of the directionof ther- dynamic compression show increases in hyaline-specific matrix apy for cartilage injury, high-quality randomized controlled synthesis [160]. Waldman and colleagues demonstrated that carti- trials with standardized outcome measures and long-term laginous tissue synthesis after intermittent compression or shear follow-up are paramount for patients and surgeons alike to force was significantly thicker (p < 0.05) and produced more discern the best treatment option. matrix (p < 0.01) than unstimulated controls [159]. Furthermore,

Expert Rev. Med. Devices 4(3), (2007) Page 23

Articular cartilage repair: procedures versus products

Paradigm of the three stages of cell-based tissue engineering

Good prognosis

I II III

Maintenance of cell Initiation of tissue Biomechanical proliferation and genesis by matrix tissue adaptation phenotypic stability production and maturation

Poor prognosis

Increasing quality of stage outcome

Product Surgeon Patient manufacturing implantation rehabilitation

Figure 7. The three stages of tissue regeneration by cell-based tissue engineering products. The outcome prognosis of cell-based therapeutic products is based on the combined success of product manufacturing, surgeon implantation of the product and maturation of the product postoperatively through appropriate patient rehabilitation. shear force produced more collagen and proteoglycan and a five- evacuated of rinsing fluid, making visualization of the procedure fold increase in equilibrium modulus when normalized against problematic. Second, measurement of the defect for graft sizing, compression cultures. These are just a few examples of the capa- while possible, is certainly nowhere near as accurate or reproduc- city to manipulate the phenotypic profile of articular chondro- ible as sizing via arthrotomy and, given the possible repair cytes for therapeutic gain. Also worth consideration are the opti- impact of poor graft-to-defect sizing, this is contraindicated. To mal number of cells for implantation, the effect of chondrocyte this end, the aforementioned development of arthroscopic com- age on clinical outcome and even whether or not to expand puter navigation for mosaicplasty is a promising avenue for the chondrocytes or chondrons given the biomechanical importance arthroscopic mapping of defect geometry in ACI [54]. Finally, of the pericellular microenvironment. While the in vitro condi- the maintenance of graft integrity during arthroscopic implant- tioning of cartilage tissue offers a promising avenue to optimize ation is also questionable. Given the importance of chondro- the treatment of articular cartilage injury by ACI, more research is cytes to the repair process, care must be taken not to compro- still needed to achieve a better mimicry of hyaline cartilage matrix mise these cells during implantation. If chondrocyte-seeded for long-lasting regenerative success. constructs are to be implanted via arthroscopic cannula, instru- While cell-based products may offer simpler surgery, more mentation must be designed to guarantee that the integrity of clinically reliable regenerative tissue and fewer complications, the graft (particularly the cells) is not affected. Given the tech- neither conventional procedures nor cell-based products can nical complexity of achieving this goal with conventional ACI claim definitive superiority in cartilage repair. However, while methodology, the most probable answer to this issue is the ACI has captured the imagination of many orthopedic research- development of 3D gel scaffolds that will allow the delivery of ers in recent years, it is limited by the requirement for arthro- chondrocytes through arthroscopic portals using minimally scopic cartilage harvest and open surgical implantation (two sur- invasive instrumentation. HA is possibly the most well publi- gical procedures), and the wait for intermediate cell expansion. cized example of this technology, although collagen gels, fibrin To this end, to facilitate faster recovery times and shorter hospi- sealant and many others exist. It is noteworthy that the develop- tal stay, cell-based cartilage repair products must evolve into a ment of photopolymerized hydrogels by Elisseeff and colleagues completely arthroscopic treatment regime. Arthroscopic MACI may prove effective for implanting encapsulated chondrocytes has been developed in Australia and Italy, but there are a few within a gel that can be polymerized in vivo during minimally issues that make this approach technically difficult. First, for invasive surgery to improve graft integration [161,162]. The bene- graft and the fibrin sealant implantation, the joint must be fits of introducing chondrocytes within a 3D gel scaffold are

www.future-drugs.com Page 24

Willers, Partsalis & Zheng many, and it has been shown to produce comparable results to grafts. As the cells are so young, they are thought to be immuno- conventional ACI [104]. Biocompatibility of the graft is suitable genically similar to harvested embryonic stem cells and, there- and no safety concerns have been reported within the literature, fore, sheltered from immune rejection. However, there is no evi- although no long-term data are currently available. Moreover, dence to confirm this theory; hence, the immunogenicity of no periosteal patch is needed and the technical difficulty of these cells remains unknown. Nevertheless, given the circumven- suturing and sealing the defect for chondrocyte introduction is tion of conventional harvest surgery, cell cultivation periods and abrogated. Therefore, the integrity of the surrounding cartilage subsequent cost savings to the patient, this technology may is not compromised by suturing. Additionally, 3D ACI can be become a new standard in cartilage repair. implemented via an arthroscopic surgical approach low in com- Although many of the current cartilage repair procedures and plexity and invasiveness, leading to decreased wound morbidity, products have demonstrated good-to-excellent clinical outcomes, hospitalization stay and surgical/recovery costs. Biologically, the the cost of these procedures to the patient varies and, hence, the delivery of chondrocytes within a 3D gel scaffold should, theo- cost–effectiveness is an inevitable consideration when evaluating retically, allow faster defect infill due to the introduced spatial the best treatment. To this end, three of the most popular carti- distribution of the cells. This may also shelter the debrided lage repair methods, mosaicplasty, MFX and ACI, have been defect borders against any postoperative deterioration, and reported to vary in their cost–effectiveness [163–166]. However, the improve integration with the freshly debrided defect walls. conclusions of these reports are only short term, subject to the But how do we remove the need for autologous cell harvest? choice of end point or based on long-term modeling and/or Research into the banking of neonatal allogenic chondrocytes or extrapolation. Hence, while comment may be made on the MSCs may allow cell-based therapy by a single procedure and short-term economics of these techniques, the dearth of pub- stabilize the increasing demand for this technology. In an inter- lished long-term data severely limits quality analysis of their esting recent development, the FDA have approved an investiga- true cost–effectiveness. While short-term symptomatic relief tional new drug application (ISTO, MO, USA) to conduct and functional restoration are one component of the development, clinical trials and global distribution of a novel car- cost–effectiveness of cartilage repair, the long-term ability of tilage repair treatment using neonatal allogenic chondrocytes. In these treatments to slow degenerative progression of the joint is, short, the technique involves taking cartilage cells from deceased perhaps more importantly, a major consideration for their true infants, then multiplying them exponentially to create many cost–effectiveness.

Key issues

• Articular cartilage injury, particular to the knee, imposes a significant socioeconomic burden on the community and shows a poor intrinsic repair ability that predicates joint degeneration. • Most conventional treatment regimens produce symptomatic joint relief but produce qualitatively and/or quantitatively poor cartilage repair tissue. • Conventional autologous chondrocyte implantation (ACI) produces good-to-excellent clinical outcomes but is plagued by high complication rates related to the use of periosteum. • The new generation of ACI using biodegradable scaffolds offers comparable clinical outcome, simplifies surgical technique and circumvents most complications but still involves two procedures. • Concomitant joint surgery may impact on the success of ACI treatment for cartilage injury of the knee to varying degrees, depending on the type and severity of the existing pathology. • Theoretically, arthroscopic allogenic chondrocyte implantation presents the most promising treatment method for articular cartilage repair, although this technology has only just entered clinical trials. • The true cost effectiveness of current cartilage repair procedures and products will only be realized with the completion of long-term randomized controlled trials, durability studies and associated economic analysis.

References 2 Clanton TO, DeLee JC. Osteochondritis 4 Aroen A, Loken S, Heir S et al. Articular Papers of special note have been highlighted as: dissecans. History, pathophysiology and cartilage lesions in 993 consecutive knee • of interest current treatment concepts. Clin. Orthop. arthroscopies. Am. J. Sports Med. 32(1), •• of considerable interest Relat. Res. 167, 50–64 (1982). 211–215 (2004). 1 Obedian RS, Grelsamer RP. 3 Curl WW, Krome J, Gordon ES, 5 Nomura E, Inoue M, Kurimura M. Osteochondritis dissecans of the distal Rushing J, Smith BP, Poehling GG. Chondral and osteochondral injuries femur and patella. Clin. Sports Med. 16(1), Cartilage injuries: a review of 31,516 knee associated with acute patellar dislocation. 157–174 (1997). arthroscopies. Arthroscopy 13(4), 456–460 Arthroscopy 19(7), 717–721 (2003). (1997).

Expert Rev. Med. Devices 4(3), (2007) Page 25

Articular cartilage repair: procedures versus products

6 Boden BP, Pearsall AW, Garrett WE Jr, 22 Jackson R. Arthroscopic Treatment of for cutting meniscal tissue: a comparative Feagin JA Jr. Patellofemoral instability: Degenerative Arthritis. McGinty JB (Ed.). experimental study on sheep menisci. evaluation and management. J. Am. Acad. Raven Press, NY, USA (1991). Arthroscopy 21(7), 844–850 (2005). Orthop. Surg. 5(1), 47–57 (1997). 23 Chang RW, Falconer J, Stulberg SD, 34 Green LM, King JS, Bianski BM, 7 Birk GT, DeLee JC. Osteochondral injuries. Arnold WJ, Manheim LM, Dyer AR. Pink MM, Jobe CM. In vitro effects of Clinical findings. Clin. Sports Med. 20(2), A randomized, controlled trial of 3 common arthroscopic instruments on 279–286 (2001). arthroscopic surgery versus closed-needle articular cartilage. Arthroscopy 22(3), 8 Bentley G, Minas T. Treating joint damage joint lavage for patients with osteoarthritis 300–307 (2006). in young people. Br. Med. J. 320(7249), of the knee. Arthritis Rheum. 36(3), 35 Steadman JR, Rodkey WG, Singleton SB, 1585–1588 (2000). 289–296 (1993). Briggs KK. Microfracture technique for 9 Elders MJ. The increasing impact of arthritis 24 Jackson RW, Marans HJ, Silver RS. The fullthickness chondral defects: technique on public health. J. Rheumatol. 60(Suppl.), arthroscopic treatment of degenerative and clinical results. Operat. Tech. Orthop. 7, 6–8 (2000). arthritis of the knee. J. Bone Joint Surg. Br. 300–304 (1997). 70, 332 (1988). 10 Yelin E. Cost of musculoskeletal diseases: 36 Steadman JR, Briggs KK, Rodrigo JJ, impact of work disability and functional 25 Moseley JB, O’Malley K, Petersen NJ Kocher MS, Gill TJ, Rodkey WG. decline. J. Rheumatol. 68, 8–11 (2003). et al. A controlled trial of arthroscopic Outcomes of microfracture for traumatic surgery for osteoarthritis of the knee. chondral defects of the knee: average 11 Ruchlin HS, Elkin EB, Paget SA. Assessing N. Engl. J. Med. 347(2), 81–88 (2002). 11-year follow-up. Arthroscopy 19(5), cost–effectiveness analyses in rheumatoid 477–484 (2003). arthritis and osteoarthritis. Arthritis Care Res. 26 Edwards RB 3rd, Lu Y, Nho S, Cole BJ, 10(6), 413–421 (1997). Markel MD. Thermal chondroplasty of 37 Frisbie DD, Oxford JT, Southwood L et al. chondromalacic human cartilage. An Early events in cartilage repair after 12 March LM, Bachmeier CJ. Economics of ex vivo comparison of bipolar and subchondral bone microfracture. Clin. osteoarthritis: a global perspective. Baillieres monopolar radiofrequency devices. Am. J. Orthop. Relat. Res. (407), 215–227 (2003). Clin. Rheumatol. 11(4), 817–834 (1997). Sports Med. 30(1), 90–97 (2002). 38 Kreuz PC, Erggelet C, Steinwachs MR 13 Brooks PM. The burden of musculoskeletal 27 Edwards RB 3rd, Lu Y, Rodriguez E, et al. Is microfracture of chondral defects in disease – a global perspective. Clin. Markel MD. Thermometric the knee associated with different results in Rheumatol. 25(6), 778–781 (2006). determination of cartilage matrix patients aged 40 years or younger? 14 Hayes DW Jr, Brower RL, John KJ. Articular temperatures during thermal Arthroscopy 22(11), 1180–1186 (2006). cartilage. Anatomy, injury, and repair. Clin. chondroplasty: comparison of bipolar and 39 Robinson D, Nevo Z. Articular cartilage Podiatr. Med. Surg. 18(1), 35–53 (2001). monopolar radiofrequency devices. chondrocytes are more advantageous for 15 Malinin T, Ouellette EA. Articular cartilage Arthroscopy 18(4), 339–346 (2002). generating hyaline-like cartilage than nutrition is mediated by subchondral bone: a 28 Hafez MI, Coombs RR, Zhou S, mesenchymal cells isolated from long-term autograft study in baboons. McCarthy ID. Ablation of bone, cartilage, microfracture repairs. Cell Tissue Bank 2(1), Osteoarthritis Cartilage 8(6), 483–491 and facet joint capsule using Ho:YAG 23–30 (2001). (2000). laser. J. Clin. Laser Med. Surg. 20(5), •Important publication in discerning the 16 Hunziker EB, Rosenberg LC. Repair of 251–255 (2002). most efficacious cell type for the cell-based partial-thickness defects in articular cartilage: 29 Janecki CJ, Perry MW, Bonati AO, treatment of cartilage injury. cell recruitment from the synovial Bendel M. Safe parameters for laser 40 Dorotka R, Windberger U, Macfelda K, membrane. J. Bone Joint Surg. Am. 78(5), chondroplasty of the knee. Lasers Surg. Bindreiter U, Toma C, Nehrer S. Repair of 721–733 (1996). Med. 23(3), 141–150 (1998). articular cartilage defects treated by 17 Dowthwaite GP, Bishop JC, Redman SN 30 Jazrawi LM, Chen A, Stein D et al. The microfracture and a three-dimensional et al. The surface of articular cartilage effects of radiofrequency bipolar thermal collagen matrix. Biomaterials 26(17), contains a progenitor cell population. J. Cell energy on human meniscal tissue. Bull. 3617–3629 (2005). Sci. 117(Pt 6), 889–897 (2004). Hosp. Joint Dis. 61(3–4), 114–117 (2003). 41 Kuo AC, Rodrigo JJ, Reddi AH, Curtiss S, • Provides the first evidence of a novel 31 Allen RT, Tasto JP, Cummings J, Grotkopp E, Chiu M. Microfracture and progenitor cell population that may Robertson CM, Amiel D. Meniscal bone morphogenetic protein 7 (BMP-7) contribute to articular cartilage debridement with an arthroscopic synergistically stimulate articular cartilage maintenance and repair. radiofrequency wand versus an repair. Osteoarthritis Cartilage 14(11), 18 Farmer JM, Martin DF, Boles CA, arthroscopic shaver: comparative effects 1126–1135 (2006). Curl WW. Chondral and osteochondral on menisci and underlying articular 42 Manolagas SC, Jilka RL. Bone marrow, injuries. Diagnosis and management. Clin. cartilage. Arthroscopy 22(4), 385–393 cytokines, and bone remodeling. Emerging Sports Med. 20(2), 299–320 (2001). (2006). insights into the pathophysiology of 19 Newman AP. Articular cartilage repair. Am. J. 32 Spahn G, Frober R, Linss W. Treatment of osteoporosis. N. Engl. J. Med. 332(5), Sports Med. 26(2), 309–324 (1998). chondral defects by hydro jet. Results of a 305–311 (1995). 20 Prakash D, Learmonth D. Natural preliminary scanning electron microscopic 43 Bonyadi M, Waldman SD, Liu D, progression of osteo-chondral defect in the evaluation. Arch. Orthop. Trauma Surg. Aubin JE, Grynpas MD, Stanford WL. femoral condyle. Knee 9(1), 7–10 (2002). 126(4), 223–227 (2006). Mesenchymal progenitor self-renewal deficiency leads to age-dependent 21 Alleyne KR, Galloway MT. Management of 33 King JS, Green LM, Bianski BM, osteoporosis in Sca-1/Ly-6A null mice. osteochondral injuries of the knee. Clin. Pink MM, Jobe CM. Shaver, bipolar Proc. Natl Acad. Sci. USA 100(10), Sports Med. 20(2), 343–364 (2001). radiofrequency, and saline jet instruments 5840–5845 (2003). www.future-drugs.com Page 26

Willers, Partsalis & Zheng

44 Oreffo RO, Bord S, Triffitt JT. Skeletal 54 Hoser C, Bichler O, Bale R et al. derived, chondrocyte-enriched cultures for progenitor cells and ageing human A computer assisted surgical technique for joint surface reconstruction. Bull. Hosp. populations. Clin. Sci. (Lond.), 94(5), retrograde autologous osteochondral Joint Dis. 53(1), 75–82 (1993). 549–555 (1998). grafting in talar osteochondritis dissecans 66 Brittberg M, Faxen E, Peterson L. Carbon 45 Hangody L, Kish G, Karpati Z, Szerb I, (OCD): a cadaveric study. Knee Surg. fiber scaffolds in the treatment of early knee Udvarhelyi I. Arthroscopic autogenous Sports Traumatol. Arthrosc. 12(1), 65–71 osteoarthritis. A prospective 4-year osteochondral mosaicplasty for the (2004). followup of 37 patients. Clin. Orthop. Relat. treatment of femoral condylar articular 55 Wu JZ, Herzog W, Hasler EM. Inadequate Res. 307, 155–164 (1994). defects. A preliminary report. Knee Surg. placement of osteochondral plugs may 67 Feczko P, Hangody L, Varga J et al. Sports Traumatol. Arthrosc. 5(4), 262–267 induce abnormal stress-strain distributions Experimental results of donor site filling for (1997). in articular cartilage – finite element autologous osteochondral mosaicplasty. 46 Bobic V. Arthroscopic osteochondral simulations. Med. Eng. Phys. 24(2), 85–97 Arthroscopy 19(7), 755–761 (2003). (2002). autograft transplantation in anterior 68 Akbar A, Diedrichs V, Wagner UA. cruciate ligament reconstruction: 56 Huang FS, Simonian PT, Norman AG, [Massive arthrofibrosis of the knee joint a preliminary clinical study. Knee Surg. Clark JM. Effects of small incongruities in after carbon fibre rod implantation. Two Sports Traumatol. Arthrosc. 3(4), 262–264 a sheep model of osteochondral cases]. Z. Orthop. Ihre Grenzgeb. 141(5), (1996). autografting. Am. J. Sports Med. 32(8), 570–572 (2003). 1842–1848 (2004). 47 Matsusue Y, Yamamuro T, Hama H. 69 Meister K, Cobb A, Bentley G. Treatment Arthroscopic multiple osteochondral 57 Kordas G, Szabo JS, Hangody L. The of painful articular cartilage defects of the transplantation to the chondral defect in the effect of drill-hole length on the primary patella by carbon-fibre implants. J. Bone knee associated with anterior cruciate stability of osteochondral grafts in Joint Surg. Br. 80(6), 965–970 (1998). ligament disruption. Arthroscopy 9(3), mosaicplasty. Orthopedics 28(4), 401–404 70 Qiu YS, Shahgaldi BF, Revell WJ, 318–321 (1993). (2005). Heatley FW. Evaluation of Gateshead 48 Kordas G, Szabo JS, Hangody L. Primary 58 Whiteside RA, Jakob RP, Wyss UP, carbon fibre rod as an implant material for stability of osteochondral grafts used in Mainil-Varlet P. Impact loading of repair of osteochondral defects: a mosaicplasty. Arthroscopy 22(4), 414–421 articular cartilage during transplantation morphological and mechanical study in the (2006). of osteochondral autograft. J. Bone Joint rabbit knee. Biomaterials 23(19), 49 Ueblacker P, Burkart A, Imhoff AB. Surg. Br. 87(9), 1285–1291 (2005). 3943–3955 (2002). Retrograde cartilage transplantation on the 59 Evans PJ, Miniaci A, Hurtig MB. Manual 71 Debnath UK, Fairclough JA, Williams RL. proximal and distal tibia. Arthroscopy 20(1), punch versus power harvesting of Long-term local effects of carbon fibre in 73–78 (2004). osteochondral grafts. Arthroscopy 20(3), the knee. Knee 11(4), 259–264 (2004). 306–310 (2004). 50 Agneskirchner JD, Brucker P, Burkart A, 72 Peterson L, Menche D, Grande D. Imhoff AB. Large osteochondral defects of 60 Huntley JS, Bush PG, McBirnie JM, Chondrocyte transplantation – an the femoral condyle: press-fit Simpson AH, Hall AC. Chondrocyte experimental rabbit in the rabbit. In: 30th transplantation of the posterior femoral death associated with human femoral Annual Orthopaedic Research Society. condyle (MEGA-OATS). Knee Surg. Sports osteochondral harvest as performed for Orthopaedic Research Society, GA, USA Traumatol. Arthrosc. 10(3), 160–168 mosaicplasty. J. Bone Joint Surg. Am. 218 (1984). (2002). 87(2), 351–360 (2005). 73 Grande DA, Pitman MI, Peterson L, 51 Karataglis D, Learmonth DJ. Management 61 Tew SR, Kwan AP, Hann A, Menche D, Klein M. The repair of of big osteochondral defects of the knee Thomson BM, Archer CW. The reactions experimentally produced defects in rabbit using osteochondral allografts with the of articular cartilage to experimental articular cartilage by autologous MEGA-OATS technique. Knee 12(5), wounding: role of apoptosis. Arthritis chondrocyte transplantation. J. Orthop. Res. 389–393 (2005). Rheum. 43(1), 215–225 (2000). 7(2), 208–218 (1989). 52 Bentley G, Biant LC, Carrington RW et al. 62 Minns RJ, Muckle DS. Mechanical and • Pioneering publication describing the A prospective, randomised comparison of histological response of carbon fibre pads potential of ACI for articular cartilage autologous chondrocyte implantation implanted in the rabbit patella. injury in the knee. versus mosaicplasty for osteochondral Biomaterials 10(4), 273–276 (1989). 74 Brittberg M, Lindahl A, Nilsson A, defects in the knee. J. Bone Joint Surg. Br. 63 Minns RJ, Flynn M. Intra-articular Ohlsson C, Isaksson O, Peterson L. 85(2), 223–230 (2003). implant of filamentous carbon fibre in the Treatment of deep cartilage defects in the •• Key publication comparing the clinical experimental animal. J. Bioeng. 2(3–4), knee with autologous chondrocyte experience of two popular treatment 279–286 (1978). transplantation. N. Engl. J. Med. 331(14), methods of articular cartilage repair, 64 Curtin W, Reville W, Heapes M, Lyons J, 889–895 (1994). mosaicplasty and autologous chondrocyte Muckle D. The chondrogenic potential of •• Key publication reporting the first clinical implantation (ACI). carbon fiber and carbon fiber periosteum outcomes of ACI for articular cartilage 53 Hangody L, Fules P. Autologous implants: an ultrastructural study in the injury in the knee. osteochondral mosaicplasty for the rabbit. Osteoarthritis Cartilage 2(4), 75 Grigolo B, Roseti L, De Franceschi L et al. treatment of full-thickness defects of 253–258 (1994). Molecular and immunohistological weight-bearing joints: ten years of 65 Robinson D, Efrat M, Mendes DG, characterization of human cartilage two experimental and clinical experience. Halperin N, Nevo Z. Implants composed years following autologous cell J. Bone Joint Surg. Am. 85-A(Suppl. 2), of carbon fiber mesh and bone-marrow- transplantation. J. Bone Joint Surg. Am. 25–32 (2003). 87(1), 46–57 (2005).

Expert Rev. Med. Devices 4(3), (2007) Page 27

Articular cartilage repair: procedures versus products

76 Peterson L, Minas T, Brittberg M, 87 Haddo O, Mahroof S, Higgs D et al. The 95 Marlovits S, Striessnig G, Nilsson A, Sjogren-Jansson E, Lindahl A. use of chondrogide membrane in Kutscha-Lissberg F et al. Early postoperative Two- to 9-year outcome after autologous autologous chondrocyte implantation. adherence of matrix-induced autologous chondrocyte transplantation of the knee. Knee 11(1), 51–55 (2004). chondrocyte implantation for the treatment Clin. Orthop. Relat. Res. 374, 212–234 88 Gooding CR, Bartlett W, Bentley G, of full-thickness cartilage defects of the (2000). Skinner JA, Carrington R, Flanagan A. femoral condyle. Knee Surg. Sports 77 Minas T, Bryant T. The role of autologous A prospective, randomised study Traumatol. Arthrosc. 13(6), 451–457 (2005). chondrocyte implantation in the comparing two techniques of autologous 96 Bartlett W, Gooding CR, Carrington RW, patellofemoral joint. Clin. Orthop. Relat. chondrocyte implantation for Skinner JA, Briggs TW, Bentley G. Res. 436, 30–39 (2005). osteochondral defects in the knee: Autologous chondrocyte implantation at the 78 Henderson IJ, Lavigne P. Periosteal periosteum covered versus type I/III knee using a bilayer collagen membrane autologous chondrocyte implantation for collagen covered. Knee 13(3), 203–210 with bone graft. A preliminary report. patellar chondral defect in patients with (2006). J. Bone Joint Surg. Br. 87(3), 330–332 normal and abnormal patellar tracking. • Important study comparing ACI using (2005). Knee 13(4), 274–279 (2006). periosteum with ACI using collagen 97 Behrens P, Bitter T, Kurz B, Russlies M. membrane for treating articular 79 Wood JJ, Malek MA, Frassica FJ et al. Matrix-associated autologous chondrocyte cartilage injury. Autologous cultured chondrocytes: adverse transplantation/implantation events reported to the United States Food 89 Bartlett W, Skinner JA, Gooding CR et al. (MACT/MACI) – 5-year follow-up. Knee and Drug Administration. J. Bone Joint Autologous chondrocyte implantation 13(3), 194–202 (2006). Surg. Am. 88(3), 503–507 (2006). versus matrix-induced autologous 98 Ehlers EM, Behrens P, Wunsch L, • Expository documentation of chondrocyte implantation for Kuhnel W, Russlies M. Effects of hyaluronic complications in a large patient cohort, osteochondral defects of the knee: a acid on the morphology and proliferation of reported following periosteal ACI prospective, randomised study. J. Bone Joint human chondrocytes in primary cell culture. treatment for articular cartilage injury. Surg. Br. 87(5), 640–645 (2005). Ann. Anat. 183(1), 13–17 (2001). • Important study comparing ACI using 80 Minas T. Autologous chondrocyte 99 Grigolo B, Roseti L, Fiorini M et al. chondrocytes injected as a suspension implantation for focal chondral defects of Transplantation of chondrocytes seeded on a against ACI using chondrocyte-seeded the knee. Clin. Orthop. Relat. Res. hyaluronan derivative (hyaff-11) into collagen membrane for treating articular 391(Suppl.), S349–S361 (2001). cartilage defects in rabbits. Biomaterials cartilage injury. 81 King PJ, Bryant T, Minas T. Autologous 22(17), 2417–2424 (2001). 90 Kirilak Y, Pavlos NJ, Willers CR et al. chondrocyte implantation for chondral 100 Kawasaki K, Ochi M, Uchio Y, Adachi N, Fibrin sealant promotes migration and defects of the knee: indications and Matsusaki M. Hyaluronic acid enhances proliferation of human articular technique. J. Knee Surg. 15(3), 177–184 proliferation and chondroitin sulfate chondrocytes: possible involvement of (2002). synthesis in cultured chondrocytes thrombin and protease-activated receptors. embedded in collagen gels. J. Cell Physiol. 82 Henderson I, Tuy B, Oakes B. Reoperation Int. J. Mol. Med. 17(4), 551–558 (2006). after autologous chondrocyte implantation. 179(2), 142–148 (1999). 91 Brittberg M, Sjogren-Jansson E, Lindahl A, Indications and findings. J. Bone Joint Surg. 101 Marcacci M, Zaffagnini S, Kon E, Visani A, Peterson L. Influence of fibrin sealant Br. 86(2), 205–211 (2004). Iacono F, Loreti I. Arthroscopic autologous (Tisseel) on osteochondral defect repair in chondrocyte transplantation: technical note. 83 Driesang IM, Hunziker EB. Delamination the rabbit knee. Biomaterials 18(3), Knee Surg. Sports Traumatol. Arthrosc. 10(3), rates of tissue flaps used in articular 235–242 (1997). cartilage repair. J. Orthop. Res. 18(6), 154–159 (2002). 92 Willers C, Chen J, Wood D, Xu J, 909–911 (2000). • Important publication demonstrating the Zheng MH. Autologous chondrocyte potential of arthroscopic ACI using a 84 Ueno T, Kagawa T, Mizukawa N, implantation with collagen bioscaffold for hyaluronic acid scaffold for articular Nakamura H, Sugahara T, Yamamoto T. the treatment of osteochondral defects in cartilage repair. Cellular origin of endochondral ossification rabbits. Tissue Eng. 11(7–8), 1065–1076 102 Gobbi A, Kon E, Berruto M, Francisco R, from grafted periosteum. Anat. Rec. 264(4), (2005). 348–357 (2001). Filardo G, Marcacci M. Patellofemoral full- 93 Zheng M, Willers C, Kirilak L et al. thickness chondral defects treated with 85 Briggs TW, Mahroof S, David LA, Matrix-induced autologous chondrocyte Hyalograft-C: a clinical, arthroscopic, and Flannelly J, Pringle J, Bayliss M. ® implantation (MACI ): biological and histologic review. Am. J. Sports Med. 34(11), Histological evaluation of chondral defects histological assessment. Tissue Eng. 13(4), 1763–1773 (2006). after autologous chondrocyte implantation 737–746 (2007). 103 Pavesio A, Abatangelo G, Borrione A et al. of the knee. J. Bone Joint Surg. Br. 85(7), • Important study assessing the histological Hyaluronan-based scaffolds (Hyalograft C) 1077–1083 (2003). progression of cartilage repair by in the treatment of knee cartilage defects: 86 Cherubino P, Grassi FA, Bulgheroni P, autologous chondrocyte implantation. preliminary clinical findings. Novartis Ronga M. Autologous chondrocyte 94 Gigante A, Bevilacqua C, Ricevuto A, Found. Symp. 249, 203–217 (2003). implantation using a bilayer collagen Mattioli-Belmonte M, Greco F. 104 Nehrer S, Domayer S, Dorotka R, membrane: a preliminary report. J. Orthop. Membrane-seeded autologous Schatz K, Bindreiter U, Kotz R. Three-year Surg. (Hong Kong) 11(1), 10–15 (2003). chondrocytes: cell viability and clinical outcome after chondrocyte • Important study describing the potential characterization at surgery. Knee Surg. transplantation using a hyaluronan matrix of replacing periosteum in ACI with Sports Traumatol. Arthrosc. 15(1), 88–92 for cartilage repair. Eur. J. Radiol. 57(1), 3–8 collagen membrane for treating articular (2007). cartilage injury. (2006). www.future-drugs.com Page 28

Willers, Partsalis & Zheng

105 Hollander AP, Dickinson SC, Sims TJ Comparison with cultured chondrocytes in 126 Im GI, Jung NH, Tae SK. Chondrogenic et al. Maturation of tissue engineered suspension. Arch. Orthop. Trauma Surg. differentiation of mesenchymal stem cells cartilage implanted in injured and 120(3–4), 121–127 (2000). isolated from patients in late adulthood: the osteoarthritic human knees. Tissue Eng. 116 Yokoyama A, Sekiya I, Miyazaki K, optimal conditions of growth factors. Tissue 12(7), 1787–1798 (2006). Ichinose S, Hata Y, Muneta T. In vitro Eng. 12(3), 527–536 (2006). 106 Grigolo B, De Franceschi L, Roseti L, cartilage formation of composites of 127 Kuroda R, Usas A, Kubo S et al. Cartilage Cattini L, Facchini A. Down regulation of synovium-derived mesenchymal stem cells repair using bone morphogenetic protein 4 degenerative cartilage molecules in with collagen gel. Cell Tissue Res. 322(2), and muscle-derived stem cells. Arthritis chondrocytes grown on a hyaluronan- 289–298 (2005). Rheum. 54(2), 433–442 (2006). based scaffold. Biomaterials 26(28), 117 Andereya S, Maus U, Gavenis K et al. [First 128 Noth U, Tuli R, Osyczka AM, 5668–5676 (2005). clinical experiences with a novel 3D- Danielson KG, Tuan RS. In vitro engineered 107 Galois L, Hutasse S, Cortial D et al. collagen gel (CaReS) for the treatment of cartilage constructs produced by press- Bovine chondrocyte behaviour in three- focal cartilage defects in the knee]. coating biodegradable polymer with human dimensional type I collagen gel in terms of Z. Orthop. Ihre Grenzgeb. 144(3), 272–280 mesenchymal stem cells. Tissue Eng. 8(1), gel contraction, proliferation and gene (2006). 131–144 (2002). expression. Biomaterials 27(1), 79–90 118 Rudert M, Wilms U, Hoberg M, Wirth CJ. 129 De Bari C, Dell’Accio F, Luyten FP. Failure of (2006). Cell-based treatment of osteochondral in vitro-differentiated mesenchymal stem 108 Chaipinyo K, Oakes BW, Van Damme MP. defects in the rabbit knee with natural and cells from the synovial membrane to form The use of debrided human articular synthetic matrices: cellular seeding ectopic stable cartilage in vivo. Arthritis cartilage for autologous chondrocyte determines the outcome. Arch. Orthop. Rheum. 50(1), 142–150 (2004). implantation: maintenance of chondrocyte Trauma Surg. 125(9), 598–608 (2005). 130 Dragoo JL, Samimi B, Zhu M et al. differentiation and proliferation in type I 119 Wakitani S, Kimura T, Hirooka A et al. Tissue-engineered cartilage and bone using collagen gels. J. Orthop. Res. 22(2), Repair of rabbit articular surfaces with stem cells from human infrapatellar fat pads. 446–455 (2004). allograft chondrocytes embedded in J. Bone Joint Surg. Br. 85(5), 740–747 109 Furthmayr H, Timpl R. collagen gel. J. Bone Joint Surg. Br. 71(1), (2003). Immunochemistry of collagens and 74–80 (1989). 131 Nathan S, Das De S, Thambyah A, Fen C, procollagens. Int. Rev. Connect Tissue Res. 120 Kim G, Okumura M, Bosnakovski D, Goh J, Lee EH. Cell-based therapy in the 7, 61–99 (1976). Ishiguro T, Kadosawa T, Fujinaga T. repair of osteochondral defects: a novel use 110 Charriere G, Bejot M, Schnitzler L, Biological properties of allogenic articular for adipose tissue. Tissue Eng. 9(4), 733–744 Ville G, Hartmann DJ. Reactions to a chondrocytes on the surface of bovine (2003). bovine collagen implant. Clinical and cartilage explants in vitro. J. Vet. Med. A 132 Schmitt B, Ringe J, Haupl T et al. BMP2 immunologic study in 705 patients. J. Am. Physiol. Pathol. Clin. Med. 50(8), 418–423 initiates chondrogenic lineage development Acad. Dermatol. 21(6), 1203–1208 (1989). (2003). of adult human mesenchymal stem cells in 111 Teixeira JO, Urist MR. Bone 121 Masuoka K, Asazuma T, Ishihara M et al. high-density culture. Differentiation morphogenetic protein induced repair of Tissue engineering of articular cartilage 71(9–10), 567–577 (2003). compartmentalized segmental diaphyseal using an allograft of cultured chondrocytes 133 Wickham MQ, Erickson GR, Gimble JM, defects. Arch. Orthop. Trauma Surg. in a membrane-sealed atelocollagen Vail TP, Guilak F. Multipotent stromal cells 117(1–2), 27–34 (1998). honeycomb-shaped scaffold (ACHMS derived from the infrapatellar fat pad of the 112 Adachi N, Ochi M, Uchio Y, Iwasa J, scaffold). J. Biomed. Mater. Res. B Appl. knee. Clin. Orthop. Relat. Res. (412), Furukawa S, Deie M. Osteochondral Biomater. 75(1), 177–184 (2005). 196–212 (2003). lesion located at the lateral femoral condyle 122 Hendrickson DA, Nixon AJ, Grande DA 134 Miao Z, Jin J, Chen L et al. Isolation of reconstructed by the transplantation of et al. Chondrocyte-fibrin matrix transplants mesenchymal stem cells from human tissue-engineered cartilage in combination for resurfacing extensive articular cartilage placenta: comparison with human bone with a periosteum with bone block: a case defects. J. Orthop. Res. 12(4), 485–497 marrow mesenchymal stem cells. Cell Biol. report. Knee Surg. Sports Traumatol. (1994). Int. 30(9), 681–687 (2006). Arthrosc. 12(5), 444–447 (2004). 123 Wakitani S, Goto T, Pineda SJ et al. 135 Zheng MH, Pembrey R, Niutta S, 113 Ochi M, Uchio Y, Kawasaki K, Mesenchymal cell-based repair of large, Stewart-Richardson P, Farrugia A. Challenges Wakitani S, Iwasa J. Transplantation of full-thickness defects of articular cartilage. in the evaluation of safety and efficacy of cartilage-like tissue made by tissue J. Bone Joint Surg. Am. 76(4), 579–592 human tissue and cell based products. ANZ J. engineering in the treatment of cartilage (1994). Surg. 76(9), 843–849 (2006). defects of the knee. J. Bone Joint Surg. Br. • Novel study looking at mesenchymal stem 136 Ochi K, Derfoul A, Tuan RS. 84(4), 571–578 (2002). cells as a source for treating articular A predominantly articular cartilage- cartilage injury. 114 Iwasa J, Ochi M, Uchio Y, Katsube K, associated gene, SCRG1, is induced by Adachi N, Kawasaki K. Effects of cell 124 Grande DA, Mason J, Light E, Dines D. glucocorticoid and stimulates chondrogenesis density on proliferation and matrix Stem cells as platforms for delivery of genes in vitro. Osteoarthritis Cartilage 14(1), 30–38 synthesis of chondrocytes embedded in to enhance cartilage repair. J. Bone Joint (2006). atelocollagen gel. Artif. Organs 27(3), Surg. Am. 85-A(Suppl. 2), 111–116 (2003). 137 Lisignoli G, Cristino S, Piacentini A, 249–255 (2003). 125 Facchini A, Lisignoli G, Cristino S et al. Cavallo C, Caplan AI, Facchini A. 115 Katsube K, Ochi M, Uchio Y et al. Repair of Human chondrocytes and mesenchymal Hyaluronan-based polymer scaffold articular cartilage defects with cultured stem cells grown onto engineered scaffold. modulates the expression of inflammatory chondrocytes in Atelocollagen gel. Biorheology 43(3–4), 471–480 (2006).

Expert Rev. Med. Devices 4(3), (2007) Page 29

Articular cartilage repair: procedures versus products

and degradative factors in mesenchymal 146 Schneider U, Anderereya S. First results of 157 Marlovits S, Tichy B, Truppe M, stem cells: involvement of Cd44 and Cd54. a prospective randomised trial of Gruber D, Schlegel W. Collagen J. Cell Physiol. 207(2), 364–373 (2006). traditional autologous chondrocyte expression in tissue engineered cartilage 138 Tatebe M, Nakamura R, Kagami H, transplantation and CaReS technology. of aged human articular chondrocytes in Okada K, Ueda M. Differentiation of Zeitschrift Orthopad. Ihre Grenzgeleit 141 a rotating bioreactor. Int. J. Artif. Organs transplanted mesenchymal stem cells in a (2003). 26(4), 319–330 (2003). large osteochondral defect in rabbit. 147 Ruano-Ravina A, Jato Diaz M. Autologous 158 Mauck RL, Wang CC, Oswald ES, Cytotherapy 7(6), 520–530 (2005). chondrocyte implantation: a systematic Ateshian GA, Hung CT. The role of cell 139 Gobbi A, Francisco RA, Lubowitz JH, review. Osteoarthritis Cartilage 14(1), seeding density and nutrient supply for Allegra F, Canata G. Osteochondral 47–51 (2006). articular cartilage tissue engineering with lesions of the talus: randomized controlled 148 Wasiak J, Clar C, Villanueva E. deformational loading. Osteoarthritis trial comparing chondroplasty, Autologous cartilage implantation for full Cartilage 11(12), 879–890 (2003). microfracture, and osteochondral thickness articular cartilage defects of the 159 Waldman SD, Spiteri CG, Grynpas MD, autograft transplantation. Arthroscopy knee. Cochrane Database Syst. Rev. 3, Pilliar RM, Hong J, Kandel RA. Effect of 22(10), 1085–1092 (2006). CD003323 (2006). biomechanical conditioning on 140 Gudas R, Stankevicius E, 149 Browne JE, Anderson AF, Arciero R et al. cartilaginous tissue formation in vitro. Monastyreckiene E, Pranys D, Clinical outcome of autologous J. Bone Joint Surg. Am. 85-A(Suppl. 2), Kalesinskas RJ. Osteochondral autologous chondrocyte implantation at 5 years in US 101–105 (2003). transplantation versus microfracture for subjects. Clin. Orthop. Relat. Res. 436, 160 Saini S, Wick TM. Concentric cylinder the treatment of articular cartilage defects 237–245 (2005). bioreactor for production of tissue in the knee joint in athletes. Knee Surg. 150 Amin AA, Bartlett W, Gooding CR et al. engineered cartilage: effect of seeding Sports Traumatol. Arthrosc. 14(9), 834–842 The use of autologous chondrocyte density and hydrodynamic loading on (2006). implantation following and combined construct development. Biotechnol. Prog. 141 Fu FH, Zurakowski D, Browne JE et al. with anterior cruciate ligament 19(2), 510–521 (2003). Autologous chondrocyte implantation reconstruction. Int. Orthop. 30(1), 48–53 161 Sharma B, Williams CG, Khan M, versus debridement for treatment of full- (2006). Manson P, Elisseeff JH. In vivo thickness chondral defects of the knee: an 151 Peterson L, Brittberg M, Kiviranta I, chondrogenesis of mesenchymal stem observational cohort study with 3-year Akerlund EL, Lindahl A. Autologous cells in a photopolymerized hydrogel. follow-up. Am. J. Sports Med. 33(11), chondrocyte transplantation. Plast. Reconstr. Surg. 119(1), 112–120 1658–1666 (2005). Biomechanics and long-term durability. (2007). 142 Knutsen G, Engebretsen L, Ludvigsen TC Am. J. Sports Med. 30(1), 2–12 (2002). 162 Lee HJ, Lee JS, Chansakul T, Yu C, et al. Autologous chondrocyte 152 Schimmer RC, Brulhart KB, Duff C, Elisseeff JH, Yu SM. Collagen mimetic implantation compared with Glinz W. Arthroscopic partial peptide-conjugated photopolymerizable microfracture in the knee. A randomized meniscectomy: a 12-year follow-up and PEG hydrogel. Biomaterials 27(30), trial. J. Bone Joint Surg. Am. 86-A(3), two-step evaluation of the long-term 5268–5276 (2006). 455–464 (2004). course. Arthroscopy 14(2), 136–142 163 Lindahl A, Brittberg M, Peterson L. •• Key study comparing the leading (1998). Health economics benefits following cartilage repair product (ACI) with the 153 Jakobsen RB, Engebretsen L, autologous chondrocyte transplantation leading cartilage repair procedure Slauterbeck JR. An analysis of the quality for patients with focal chondral lesions of (microfracture) in the treatment of of cartilage repair studies. J. Bone Joint the knee. Knee Surg. Sports Traumatol. articular cartilage injury. Surg. Am. 87(10), 2232–2239 (2005). Arthrosc. 9(6), 358–363 (2001). 143 Basad E, Stürz H, Steinmeyer J. •• Pertinent commentary on the quality of 164 Clar C, Cummins E, McIntyre L et al. [Treatment of chondral defects with current cartilage repair studies and areas Clinical and cost–effectiveness of MACI or microfracture. First results of a for improvement. autologous chondrocyte implantation for comparative clinical study.] Orthopädische 154 Lu Y, Adkisson HD, Bogdanske J et al. cartilage defects in knee joints: systematic Praxis 40, 6–10 (2004). In vivo transplantation of neonatal ovine review and economic evaluation. Health 144 Dozin B, Malpeli M, Cancedda R et al. neocartilage allografts: determining the Technol. Assess. 9(47), iii–iv, ix–x, 1–82 Comparative evaluation of autologous effectiveness of tissue transglutaminase. (2005). chondrocyte implantation and J. Knee Surg. 18(1), 31–42 (2005). 165 Derrett S, Stokes EA, James M, mosaicplasty: a multicentered randomized 155 Feder J, Adkisson HD, Kizer N et al. The Bartlett W, Bentley G. Cost and health clinical trial. Clin. J. Sport Med. 15(4), promise of chondral repair using status analysis after autologous 220–226 (2005). neocartilage. In: Tissue Engineering in chondrocyte implantation and 145 Fu FH. Rate of improvement was not Musculoskeletal Clinical Practice. Sandell L, mosaicplasty: a retrospective comparison. different after osteochondral repair with Grodzinsky A (Eds). American Academy Int. J. Technol. Assess. Health Care 21(3), matrix-induced autologous chondrocyte of Orthopaedic Surgeons, PA, USA, 359–367 (2005). implantation or autologous chondrocyte 219–226 (2004). 166 Wildner M, Sangha O, Behrend C. implantation with a cover made from 156 Weinand C, Peretti GM, Adams SB Jr, Wirtschaftlichkeitsuntersuchung zur porcine-derived type I/type III collagen. Bonassar LJ, Randolph MA, Gill TJ. autologen Chondrozytentransplantation. J. Bone Joint Surg. Am. 87(11), 2593 An allogenic cell-based implant for Arthroskopie 13, 123–131 (2000). (2005). meniscal lesions. Am. J. Sports Med. 34(11), 1779–1789 (2006). www.future-drugs.com Page 30

Willers, Partsalis & Zheng

Website • Theo Partsalis, MBBS • Ming-Hao Zheng, PhD, DM, FRCPath Department of Orthopaedics, Professor, Director of Research, 201 TIGENIX. TiGenix announces positive School of Pathology and Surgery, Department of Orthopaedics, Phase III trial results. Beyen G, Motmans K University of Western Australia, School of Pathology and Surgery, (Eds). 2nd Floor, M-block, QEII Medical Centre, University of Western Australia, www.tigenix.com Nedlands, Perth, WA 6009, Australia 2nd Floor, M-block, QEII Medical Centre, Tel.: +61 089 346 3213 Nedlands, Perth, WA 6009, Australia Affiliations Fax: +61 089 346 3210 Tel.: +61 089 346 3213 •Craig Willers, BSc(H1), M(Med)Sc [email protected] Fax: +61 089 346 3210 Department of Orthopaedics, [email protected] School of Pathology and Surgery, University of Western Australia, 2nd Floor, M-block, QEII Medical Centre, Nedlands, Perth, WA 6009, Australia Tel.: +61 089 346 3213 Fax: +61 089 346 3210 [email protected]

Expert Rev. Med. Devices 4(3), (2007)

Chapter 3

Molecular Biology of the MACI Technique

Page 31

STATEMENTS OF AUTHORSHIP CONTRIBUTION FROM CO-AUTHORS

Fibrin sealant promotes migration and proliferation of human articular chondrocytes: possible involvement of thrombin and protease-activated receptors.

Kirilak Y, Pavlos NJ, Willers CR, Han R, Feng H, Xu J, Asokananthan N, Stewart GA, Henry P, Wood D, Zheng MH.

Published in the International Journal of Molecular Medicine 2006, volume 17(4), pages 551-558.

Kirilak Y (Research Collaborator) Executed most research including cell culture, MACI co-culture, RT-PCR, and immunohistochemistry. Moderate contribution to planning of research. Moderate contribution to analyzing research data.

Signature of Principal Author :……………………………… Date: .………………

Pavlos N (Co-Supervisor) Executed confocal microscopy. Moderate contribution to planning of research. Moderate contribution to analyzing and interpreting research data. Minor contribution to manuscript writing.

Signature of Co-Author: ……………………..……………… Date: …………………

Willers C (PhD Candidate) Executed histology and microscopy. Minor contribution to planning of research. Moderate contribution to interpreting research data. Major contribution to manuscript writing.

Signature of Co-Author: ……………………..……………… Date: ………………… Page 32

Han R (Research Collaborator) Executed BrdU proliferation assay. Minor contribution to analysing research data. Minor contribution to manuscript writing.

Signature of Co-Author:……………………..………………Date:…………………

Feng H (Research Collaborator) Executed confocal microscopy. Minor contribution to analysing research data.

Signature of Co-Author:……………………..………………Date:…………………

Xu J (Research Collaborator) Assisted with cell culture and RT-PCR. Reviewed manuscript prior to submission.

Signature of Co-Author:……………………..………………Date:…………………

Asokananthan N (Research Collaborator) Conducted intracellular calcium assay.

Signature of Co-Author:……………………..………………Date:…………………

Stewart G.A. (Research Collaborator) Assisted with intracellular calcium assay. Reviewed manuscript prior to submission.

Signature of Co-Author:……………………..………………Date:…………………

Henry P (Research Collaborator) Moderate planning of project. Moderate analysis of data.

Signature of Co-Author:……………………..………………Date:………………… Page 33

Wood D (Co-Supervisor) Provided cartilage for cell culture. Reviewed manuscript prior to submission.

Signature of Co-Author:……………………..………………Date:…………………

Zheng M.H. (Supervisor) Major planning of project. Supervised MACI co-culture. Supervised interpretation of data. Reviewed manuscript prior to submission.

Signature of Co-Author:……………………..………………Date:………………… Page 34

STATEMENTS OF PERMISSION FROM CO-AUTHORS

I, Yaowanuj Kirilak give permission to Craig Willers, third author of the paper: Fibrin sealant promotes migration and proliferation of human articular chondrocytes: possible involvement of thrombin and protease-activated receptors., published in the International Journal of Molecular Medicine (2006), to include this paper as Chapter 3 in his PhD thesis Matrix-induced Autologous Chondrocyte Implantation for Articular Cartilage Injury: Biology, Histology, and Clinical Outcomes.

Signature: ………………………………… Date: ………………………………….

I, Nathan Pavlos give permission to Craig Willers, third author of the paper: Fibrin sealant promotes migration and proliferation of human articular chondrocytes: possible involvement of thrombin and protease-activated receptors., published in the International Journal of Molecular Medicine (2006), to include this paper as Chapter 3 of his PhD thesis entitled Matrix-induced Autologous Chondrocyte Implantation for Articular Cartilage Injury: Biology, Histology, and Clinical Outcomes.

Signature: ………………………………… Date: ………………………………….

I, Renzhi Han give permission to Craig Willers, third author of the paper: Fibrin sealant promotes migration and proliferation of human articular chondrocytes: possible involvement of thrombin and protease-activated receptors., published in the International Journal of Molecular Medicine (2006), to include this paper as Chapter 3 of his PhD thesis entitled Matrix-induced Autologous Chondrocyte Implantation for Articular Cartilage Injury: Biology, Histology, and Clinical Outcomes.

Signature: …………………………………. Date: ………………………………….

Page 35

I, Haotian Feng give permission to Craig Willers, third author of the paper: Fibrin sealant promotes migration and proliferation of human articular chondrocytes: possible involvement of thrombin and protease-activated receptors., published in the International Journal of Molecular Medicine (2006), to include this paper as Chapter 3 of his PhD thesis entitled Matrix-induced Autologous Chondrocyte Implantation for Articular Cartilage Injury: Biology, Histology, and Clinical Outcomes.

Signature: ………………………………… Date: ………………………………….

I, Jiake Xu give permission to Craig Willers, third author of the paper: Fibrin sealant promotes migration and proliferation of human articular chondrocytes: possible involvement of thrombin and protease-activated receptors., published in the International Journal of Molecular Medicine (2006), to include this paper as Chapter 3 of his PhD thesis entitled Matrix-induced Autologous Chondrocyte Implantation for Articular Cartilage Injury: Biology, Histology, and Clinical Outcomes.

Signature: ………………………………… Date: ………………………………….

I, Nithiananthan Asokananthan give permission to Craig Willers, third author of the paper: Fibrin sealant promotes migration and proliferation of human articular chondrocytes: possible involvement of thrombin and protease-activated receptors., published in the International Journal of Molecular Medicine (2006), to include this paper as Chapter 3 of his PhD thesis entitled Matrix-induced Autologous Chondrocyte Implantation for Articular Cartilage Injury: Biology, Histology, and Clinical Outcomes.

Signature: ………………………………… Date: ………………………………….

Page 36

I, Geoffrey Stewart give permission to Craig Willers, third author of the paper: Fibrin sealant promotes migration and proliferation of human articular chondrocytes: possible involvement of thrombin and protease-activated receptors., published in the International Journal of Molecular Medicine (2006), to include this paper as Chapter 3 of his PhD thesis entitled Matrix-induced Autologous Chondrocyte Implantation for Articular Cartilage Injury: Biology, Histology, and Clinical Outcomes.

Signature: …………………………………. Date: ………………………………….

I, Peter Henry give permission to Craig Willers, third author of the paper: Fibrin sealant promotes migration and proliferation of human articular chondrocytes: possible involvement of thrombin and protease-activated receptors., published in the International Journal of Molecular Medicine (2006), to include this paper as Chapter 3 of his PhD thesis entitled Matrix-induced Autologous Chondrocyte Implantation for Articular Cartilage Injury: Biology, Histology, and Clinical Outcomes.

Signature: …………………………………. Date: ………………………….

I, David Wood give permission to Craig Willers, third author of the paper: Fibrin sealant promotes migration and proliferation of human articular chondrocytes: possible involvement of thrombin and protease-activated receptors., published in the International Journal of Molecular Medicine (2006), to include this paper as Chapter 3 of his PhD thesis entitled Matrix-induced Autologous Chondrocyte Implantation for Articular Cartilage Injury: Biology, Histology, and Clinical Outcomes.

Signature: …………………………………. Date: ………………………………….

Page 37

I, Ming-Hao Zheng give permission to Craig Willers, third author of the paper: Fibrin sealant promotes migration and proliferation of human articular chondrocytes: possible involvement of thrombin and protease-activated receptors., published in the International Journal of Molecular Medicine (2006), to include this paper as Chapter 3 of his PhD thesis entitled Matrix-induced Autologous Chondrocyte Implantation for Articular Cartilage Injury: Biology, Histology, and Clinical Outcomes.

Signature: …………………………………. Date: ………………………………….

Page 38

INTERNATIONAL JOURNAL OF MOLECULAR MEDICINE 17: 551-558, 2006 551

Fibrin sealant promotes migration and proliferation of human articular chondrocytes: Possible involvement of thrombin and protease-activated receptors

YAOWANUJ KIRILAK1*, NATHAN J. PAVLOS1*, CRAIG R. WILLERS1, RENZHI HAN1, HAOTIAN FENG1, JIAKE XU1, NITHIANANTHAN ASOKANANTHAN2, GEOFFREY A. STEWART2,4, PETER HENRY3, DAVID WOOD1 and MING H. ZHENG1

1School of Surgery and Pathology (Orthopaedics), 2School of Biomedical and Chemical Sciences (Microbiology), 3School of Medicine and Pharmacology (Pharmacology), The University of Western Australia, 35 Stirling Highway, Crawley WA 6009; 4Western Australian Institute for Medical Research, Australia

Received October 31, 2005; Accepted December 5, 2005

Abstract. Fibrin sealant (FS), a biological adhesive material, a PAR-1, but not other PAR-isotype-specific peptide agonists, has been recently recommended as an adjunct in autologous were found to induce rapid intracellular Ca2+ responses in chondrocyte implantation (ACI). While FS has been shown to human chondrocytes in calcium mobilization assays. Together, possess osteoinductive potential, little is known about its these data demonstrate that FS supports both the migration effects on chondrogenic cells. In this study, we assessed the and proliferation of human chondrocytes. We propose that bioactivity of FS (Tisseel®) on the migration and proliferation these effects are mediated, at least in part, via thrombin- of human articular chondrocytes in vitro. Using a co-culture induced PAR-1 signalling in human chondrocytes. assay to mimic matrix-induced ACI (MACI), chondrocytes were found to migrate from collagen membranes towards FS Introduction within 12 h of culture, with significant migratory activity evident by 24 h. In addition, 5-bromo-2'-deoxyuridine Fibrin sealants (FSs) have long been utilised as an adjunct (BrdU) incorporation experiments revealed that thrombin, the in a variety of surgical procedures to promote hemostasis and active component of the tissue glue, stimulated chondrocyte tissue sealing (1-4). In orthopaedics, FSs are used as a tissue proliferation, with maximal efficacy observed at 48 h post- adhesive for the fixation of osteochondral fragments and stimulation (1-10 U/ml). In an effort to elucidate the molecular fractures (5), in spinal surgery (6), and securing perichondral mechanisms underlying these thrombin-induced effects, we grafts (7). More recently, fibrin preparations have been examined the expression and activation of protease-activated employed as biological vehicles for delivering chondrocytes receptors (PARs), established thrombin receptors. Using a directly to cartilage defects in order to stimulate repair combination of RT-PCR and immunohistochemistry, all four processes, however with contradicting outcomes reported in PARs were detected in human chondrocytes, with PAR-1 the literature. Studies by Homminga et al demonstrated that being the major isoform expressed. Moreover, thrombin and chondrocytes encapsulated in FS retained their morphology and actively synthesised matrix suggesting that the adhesive served an effective matrice (8). Similarly, Hendrickson et al ______reported that FS-bound allogenic chondrocyte grafts displayed significantly higher glycosaminoglycan and collagen II content 8 months post-implantation (9). On the other hand, Correspondence to: Professor Ming-Hao Zheng, Unit of in vivo studies by Brittiberg et al reported opposite effects Orthopaedics, School of Surgery and Pathology, University of Western Australia, 2nd Floor M-block QEII Medical Centre, suggesting that FS-derived scaffolds were not suitable for Nedlands 6009, Australia osteochondrocal healing (10). While ambiguity remains over E-mail: [email protected] the application of FSs in articular cartilage repair, recent studies indicate that these biological adhesives possess *Contributed equally unique osteoinductive properties (11). Furthermore, there is increasing evidence to suggest that FSs support the growth Key words: fibrin sealant, thrombin, autologous chondrocyte and migration of chondrocytes (12-15). Based on the collation implantation, calcium mobilization, collagen membrane, protease- of these and recent studies, the use of FSs as a component of activated receptors autologous chondrocyte implantation (ACI) has now been advocated. Thrombin, a coagulative serine protease, is an active ingredient of FSs. Thrombin is ubiquitously expressed at Page 39

KIRILAK et al: EFFECTS OF FIBRIN SEALANT ON HUMAN CHONDROCYTES

sites of vascular injury where it serves to accelerate the fetal calf serum (FCS) (Gibco), 292 μg/ml L-glutamine, coagulation process via proteolytic cleavage of fibrinogen 10,000 U/ml penicillin G, 10,000 U/ml streptomycin sulfate, (16). In addition to its role in wound healing, thrombin has 25 μg/ml amphotericin B, and 50 μg/ml ascorbic acid. Cells been shown to induce a variety of cellular responses including were sub-cultured routinely as required per experimental proliferation (17-21), migration (22-25) and survival (26,27). condition. These diverse biological effects are mediated through specific interaction(s) with cell surface receptors. Among the candidate In vitro MACI co-culture assay. To simulate the in vivo thrombin receptors, members of the seven transmembrane G conditions following matrix-induced autologous chondrocyte protein-coupled protease activated-receptor (PAR) family implantation (MACI) (12), primary human chondrocytes (PAR-1, PAR-2, PAR-3, PAR-4) are perhaps the best (5.0x106 cells/ml) were seeded onto a 1.5 cm2 type I/III characterised (28). Thrombin is known to activate intracellular collagen membrane (Matricel®, Germany) in 6-well plates signalling of PAR-1, 3 and 4 via cleavage of the extracellular and left to attach for at least 24 h. Following attachment, N-terminal domain, which unmasks a ‘tethered ligand’ Tisseel FS (containing 500 U/ml thrombin) was applied to sequence which binds intramolecularly to a receptor domain the cell-seeded surface of the membrane scaffold using the thereby activating G protein-coupled signal transduction supplied duojector. The resulting membrane-cell-FS pathways (29,30). By comparison, PAR-2 is activated by ‘sandwiches’ were then cultured in complete growth media trypsin and tryptase-associated proteases, but not thrombin for 12, 24, 48 h, and 15 days to promote cell migration. Seeded (31-33). PAR-1, -2 and -4 can also be activated without collagen I/III membranes in the absence of FS served as proteolytic cleavage, using five to six amino-acid residue controls. Migration patterns of chondrocytes towards the FS peptides corresponding to the new amino termini of the were scored either histologically or by phase-contrast light cleaved receptors (28). microscopy (Nikon Diaphot). The goal of this study was to assess the bioactive properties of commercial FS (Tisseel®), with particular emphasis on the Histology. MACI-FS ‘sandwiches’ were carefully removed thrombin component, on autologous human chondrocyte from culture medium and washed twice in 1X PBS before migration and proliferation in vitro. In addition, we examined being fixed in ice-cold paraformaldehyde (4%; 15 min) at room the expression and localisation of PARs in cultured human temperature. Following fixation, the ‘sandwiches’ were washed chondrocytes. Our findings indicate that FS induces strong 3x with 1X PBS before undergoing routine paraffin processing chemotactic and mitogenic responses in cultured chondrocytes. and embedding. All samples were embedded in vertical Furthermore, we provide evidence to suggest that these effects orientation to the cutting plane so that both surfaces of the are mediated, at least in part, via thrombin-induced activation ‘sandwiches’ were displayed during tissue sectioning. of PAR-1 signalling. Sections were cut (4-6 μm), placed onto glass slides and then de-waxed (xylene: 2x 3-4 min; 100% ethanol: 2x 3-4 min; Materials and methods 95% and 70% ethanol: 1x 3 min each). All sections were stained with Gill's hematoxylin and eosin, mounted with Materials. Tisseel FS was purchased from Baxter AG (Vienna, Depex, and examined by light microscopy. Austria). Tissue culture reagents and molecular biology reagents were purchased from Life Technologies (Melbourne, Proliferation assay. Cell proliferation was assayed by 5-bromo- Australia) and Stratagene (La Jolla, CA, USA) respectively. 2'-deoxyuridine (BrdU) using a commercial available Biotrak The Biotrak® cell proliferation ELISA system was purchased cell proliferation ELISA system. Briefly, human chondrocytes from Amersham Life Sciences (Buckinghamshire, UK). (5x103 cells/well) were cultured in 96-well plates in complete Synthetic agonist PAR peptides with amidated C termini growth medium overnight. Following attachment, cells were

(PAR-1, SFLLRN-NH2, TFLLRN-NH2; PAR-2, SLIGKV- washed twice with 1X phosphate-buffered-saline (PBS) and

NH2; PAR-3, TFRGAP-NH2; and PAR-4, GYPGQV-NH2; deprived of serum for an additional 24 h before the addition purity >85%) were synthesized by the Protein Facility, of thrombin (0.1, 0.5, 1, 10 U/ml) or vehicle (CaCl2, 40 mM) in University of Western Australia, Perth, Australia. All other combination with serum-free DMEM F-12 for an additional chemicals were purchased from Sigma-Aldrich (St. Louis, 24-48 h. BrdU-labeling solution was then added to each well, MO, USA) unless stated otherwise. and cells were re-incubated for an additional 16 h. Time points (24 and 48 h) were staggered so that addition of BrdU occurred Cell culture. Cartilage biopsies obtained from healthy human on the same day. Cells were fixed, and incorporated BrdU patients were used as a source of chondrocytes. All patients was detected using immunoperoxidase and tetramethyl- consented and ethics approval was obtained through the benzidine (TMB) according to the manufacturer’s protocol. University of Western Australia Human Research Ethics Absorbance was read at 450 nm. Committee. In brief, biopsies were mechanically disaggregated and then digested with 0.3% (w/v) collagenase type II Reverse transcription (RT)-PCR. Total RNA was extracted (Worthington Biochemical Corp., NJ, USA) at 37˚C for 5-8 h from the primary human chondrocytes using RNAzol B with shaking to release chondrocytes. Following digestion, according to the manufacturer’s instructions (Tel-test, TX). resulting chondrocyte suspensions were passed through a cDNA was synthesised from 2 μg of total RNA using the 100 μm cell strainer (Becton Dickinson and Co., NJ, USA) RETROscript™ First-strand synthesis kit (Ambion). Primers before being cultured in a T75-culture flask containing DMEM against human PAR isoforms were designed based on F-12 media (Gibco, NY, USA) supplemented with 10% (v/v) published sequence data (34) and purchased from Genset Page 40

INTERNATIONAL JOURNAL OF MOLECULAR MEDICINE 17: 551-558, 2006

Pacific Pty. Ltd., Australia. PAR-1, sense 5'-TGTGAACTGA processed using Confocal Assistant 4.02. All images were TCATGTTTATG-3', anti-sense 5'-TTCGTAAGATAAGAG collected under non-saturating conditions set up by the use of ATATGT-3' (PCR product, 708 bp); PAR-2, sense 5'-AGAA an output look-up table (LUT). GCCTTATTGGTAAGGTT-3', anti-sense 5'-AACATCATG ACAGGTCGTGAT-3' (PCR product, 582 bp); PAR-3, sense Intracellular Ca2+ mobilization. Intracellular calcium 5'-CTGATACCTGCCATCTACCTCC-3, anti-sense 5'-AG mobilization was measured fluorimetrically using Fura-2/AM AAAACTGTTGCCCACACC-3' (PCR product, 382 bp); (Molecular Probes, Eugene, OR, USA). Cells were trypsinised PAR-4, sense 5'-ATTACTCGGACCCGAGCC-3', anti-sense and seeded onto 10-mm coverslips in 35-mm culture dishes. 5'-TGTAAGGCCCACCCTTCTC-3' (PCR product, 392 bp). Upon reaching confluence, cells were washed twice with Amplification of ß-actin, with the sense and anti-sense primer freshly prepared physiological rodent saline (PRS, 138 mM pair 5'-GGCTCTTCCAGCCTTCCTTCCT-3' and 5'-CACA NaCl, 2.7 mM KCl, 1.8 mM CaCl2, 1.06 mM MgCl2, 12.4 mM GAGTACTTGCGCTCAGGAGG-3' (PCR product 240 bp), HEPES, 5.6 mM glucose, and 1 mM probenecid; pH 7.3) and served as an internal loading control. PCRs were performed in loaded for 45 min at (37˚C) with Fura-2/AM (3 μM) and 25 μl reaction volumes containing 2 μl of cDNA, 1 μl of 5 mM 0.0125% F-127 (w/v) in PRS (1x105 cells/ml). Following dNTPs, 2.5 μl of 10X buffer (Boehringer Mannheim, loading, cells were washed and incubated in the dark (30 min, Germany), 0.2 μl of Taq DNA polymerase (Boehringer room temperature). Coverslips were then removed and Mannheim) and water to 25 μl. Amplifications were performed carefully placed into a heated bio-chamber (37˚C) attached to in a DNA thermal cycler (model 2400; Perkin-Elmer). Cycling an inverted epifluorescence microscope (Nikon TE2000, parameters were 94˚C for 3 min; 35 cycles: 94˚C for 40 sec, Japan). Fluorescence emission (510 nm) at 340- and 380-nm annealing at 55˚C (PAR-1 and GAPDH), 60˚C (PAR-2) or excitation was measured using a spectrophotometer (Cairn, 64˚C (PAR-3 and PAR-4) for 40 sec, extension 72˚C for 40 sec; UK). Thrombin (1 U/ml) and PAR-1, 3 and 4 agonists and a final extension 72˚C for 10 min. PCR products were (400 μM) were added into the chamber following 1 min resolved on a 1.5% (w/v) agarose gel containing ethidium background recording and corresponding fluorescence bromide and visualised and photographed on a UV trans- emission ratios were recorded for 5 min. The PAR-2 agonist illuminator. was not assessed because of its known activation by trypsin and tryptase (31-33). All intracellular calcium concentrations Immunocytochemistry. Immunodetection of PARs was were expressed as the ratio of emission following excitation conducted on human chondrocytes cultured in 8-well chamber of 340 and 380 nm respectively. slides (LAB-Tek II; Nunc) according to methods previously described by Asokananthan et al (35). Briefly, upon reaching Statistical analyses. Unless stated otherwise, all data are approximately 80% confluence, cells were washed twice with expressed as mean ± SEM. Statistical significance between PBS before being fixed with 4% (v/v) paraformaldehyde. means was determined by ANOVA or the Student's t-test using Endogenous peroxidase activity was quenched by incubating GraphPad Prism (GraphPad Software, San Diego, CA, USA). chamber slides in 3% (v/v) H2O2 for 5 min, and non-specific P-values <0.01 were considered significant. binding was blocked by incubation in 10% FCS in PBS for 1 h. Cells were then incubated with primary antibodies raised Results against specific PAR isoforms: mouse monoclonal anti-human PAR-1 (ATAP2: sc-13503, Santa Cruz), mouse monoclonal Fibrin sealant promotes chondrocyte migration in vitro. Our anti-human PAR-2 (SAM11: sc-13504, Santa Cruz), rabbit previous in vivo studies indicate that FS (Tisseel®) stimulates polyclonal anti-PAR-3 (raised against peptide 37TLPIKT migration of autologous chondrocytes to osteochondral defects FRGAPPNSFEEFP55; and rabbit polyclonal anti-PAR-4 suggesting that the sealant possesses chemoattractive properties

(raised against peptide 28EDDSTPSLLPAPRGYPGQV39) (35). (15). In order to determine whether the observed chondrocytic Following the addition of secondary antibodies (either bio- migration was directly related to chemotactic activity of FS, or tinylated anti-mouse or anti-rabbit IgG), PAR expression was a subsidiary effect of the repair process in vivo, we sought to visualised using streptavidin peroxidase and diaminobenzidine replicate MACI using an in vitro co-culture system. For this (DAB). Incubation with either pre-immune serum or staining purpose, autologous human chondrocytes, grown on a type I/III for the chondrocytic marker S-100 (rabbit polyclonal 18-0046, collagen membrane (Matricel), were ‘sandwiched’ between Zymed Laboratories Inc.) served as negative and positive FS and cultured in vitro for 12-48 h. Following incubation, the controls respectively. ‘cell sandwiches’ were fixed and the migratory activities Immunolocalisation studies were performed essentially as assessed histologically. As shown in Fig. 1A, chondrocytic previously outlined in Pavlos et al (36) using secondary anti- migration from the collagen scaffold toward the FS was bodies (dilution of 1:1000): goat anti-rabbit or goat anti-mouse observed as early as 12 h co-culture, although the majority of immunoglobulin G conjugated to Alexa Fluor 488 (Molecular cells remained on the superficial surface of the collagen Probes Inc.). Cell nuclei were visualised by counter-staining membrane, reminiscent to that of the control. At 24 and 48 h with Hoechst 33342 (1:10,000) (Molecular Probes Inc.). time points, substantial cell migration was evident, with no Detection of fluorochromes was carried out by confocal laser breaching of the FS surface observed. We also assessed scanning microscopy (CLSM) (MRC-1000, Bio-Rad), chondrocytic migration in co-cultures incubated for up to equipped with a krypton-argon laser or argon ion laser coupled 15 days. As shown in Fig. 1B, extensive migration of to an epifluorescence Nikon Diaphot 300 inverted microscope. chondrocytes towards the FS-collagen interface is evident, Confocal sequences were collected as Bio-Rad PIC files and with some cells clearly invading the fibrin matrix. Collectively, Page 41

KIRILAK et al: EFFECTS OF FIBRIN SEALANT ON HUMAN CHONDROCYTES

Figure 2. Proliferative response of human articular chondrocytes to different concentrations of thrombin. Chondrocytes were seeded onto 96-well plates, serum-starved for 24 h and then stimulated with varying concentrations of

thrombin (0.1-10 U/ml) for an additional 24-48 h. Vehicle (CaCl2, 40 mM) served as a control. BrdU-incorporation as a parameter of proliferation was measured, and expressed as absorbance at 450 nm. Data are expressed as mean ± SEM (n=6). *p<0.005; **p<0.0005.

Figure 3. mRNA expression of PARs in human articular chondrocytes. Total RNA was isolated from cultured primary human chondrocytes and subjected Figure 1. (A) Histology of chondrocyte migration from collagen membrane to RT-PCR using PAR-isoform-specific primers. Reaction products were to FS from 12 to 48 h. Chondrocyte migration from the collagen membrane resolved by agarose gel (1.5%) electrophoresis, stained with ethidium towards the fibrin surface was seen after a 12-h culture (arrows). At 24 and bromide, and visualized under a UV transilluminator. PAR-1 (708 bp) and 48 h, aggregation of chondrocytes on the surface of the FS is clearly PAR-3 (382 bp) are highly expressed in chondrocytes whereas PAR-2 evident. In contrast, chondrocyte-seeded collagen membrane alone at 12, 24, (582 bp) and PAR-4 (392 bp) exhibit moderate to weak expression and 48 h showed in-growth of chondrocytes into the collagen matrix of the respectively. ß-actin (240 bp) served as an internal loading control. membrane. All sections were stained with H&E. Magnification, x25. (B) Chondrocyte migration and in-growth into FS at 15 days under phase- contrast microscopy. As seen at both x25 and x40 magnifications, chondrocyte in-growth was visualized within the FS (arrows). AC, (0.1-10 U/ml) for 24 and 48 h time points and its effect on autologous chondrocytes; CM, collagen membrane; FS, fibrin sealant. cell proliferation was assessed using an ELISA-based BrdU incorporation assay. As shown in Fig. 2, thrombin induced a marked increase in BrdU incorporation with maximal response observed at doses of 10 U/ml (24 h: p<0.005; 48 h: p<0.0005). these data corroborate the notion that FS promotes the Significant proliferative activity was also detected in cells migration of autologous chondrocytes. cultured in the presence of 1 U/ml thrombin, however only for the 48 h time point (p<0.005). On the other hand, lower Thrombin stimulates proliferation of human chondrocytes. concentrations of thrombin (<1 U/ml) failed to elicit any Having established that FS stimulates chondrocytic migration significant proliferative response. Comparable results were we next asked the question whether the sealant also possessed obtained using Alamar blue cell proliferation assays (data not mitogenic potential. Given that thrombin, the active constituent shown). In all, these data indicate that the thrombin component of FS, has been previously shown to induce proliferation of a of FS supports chondrocyte proliferation in vitro. variety of cell types including endothelial cells (37), neutrophils (38), and osteoblasts (20), we examined whether the thrombin PAR-1 is the major PAR isoform expressed in human chon- component similarly influenced the proliferative capacity of drocytes. Numerous studies indicate that thrombin elicits its articular chondrocytes. To this end, human chondrocytes biological responses via its interaction with, and subsequent were incubated with increasing concentrations of thrombin activation of, PARs (28). Therefore, as an initial step towards Page 42

INTERNATIONAL JOURNAL OF MOLECULAR MEDICINE 17: 551-558, 2006

Figure 5. Subcellular localisation of PARs in human chondrocytes. Chondrocytes were grown on 8-well chamber slides, fixed with 4% paraformaldehyde, and then immunostained for PAR isoforms (green). All slides were counter-stained with Hoescht 33342 to visualise nuclei (blue) and images were recorded by confocal microscopy. Marked staining of the plasma membrane and cytosol was detected for PAR-1, -3 and -4. PAR-2 also shows plasma membrane staining as well as a sub-population of perinuclear vesicles (inset). Bar = 10 μm.

Figure 4. Immunocytochemical detection of PARs in cultured human chondrocytes. Strong expression of both PAR-1 and PAR-3 is evident in human chondrocytes. PAR-2 displayed moderate chondrocytic expression whilst negligible PAR-4 staining was detectable. Immunostaining for S-100 served as a positive marker for chondrocytes. Sections were viewed at x250 and x400 magnification.

elucidating the molecular mechanism(s) underlying thrombin- Figure 6. Intracellular calcium responses in human chondrocytes elicited by induced chondrocyte proliferation, we examined the mRNA thrombin and PAR agonist peptides. Human chondrocytes were loaded with expression of PARs in chondrocytes by RT-PCR using Fura-2/AM and incubated either with thrombin (1 U/ml) or PAR-1, -3, and -4 agonist peptides at 400 μM. Thrombin (A) and PAR-1 agonist peptide isoform-specific primers. As shown in Fig. 3, PAR-1 and -3 SFLLRN-NH2 (B) induced steep intracellular calcium elevations upon mRNAs were highly expressed in cultured human chondro- addition, but thrombin elevation did not return to baseline. All responses cytes. By comparison, PAR-2 and PAR-4 exhibited moderate were measured over a 5-min period and results were expressed as a to weak expression respectively. fluorescence ratio (340/380 nm). Traces are representative of at least 5 To confirm the PCR data, we next assessed PAR protein independent experiments. expression levels by immunocytochemistry using a number of well-characterised PAR isoform-specific antisera (35) (Fig. 4). Consistent with the observed mRNA expression not shown). We also examined the subcellular localization of patterns, both PAR-1 and PAR-3 were strongly expressed in the PAR isoforms by confocal microscopy (Fig. 5). Whereas human chondrocytes, with PAR-1 exhibiting a slightly higher PAR-1, -3 and -4 isoforms were largely localized to the plasma level of expression. On the other hand, PAR-2 displayed membranes, with diffuse/reticular-like staining throughout the moderate staining whereas PAR-4 expression was detectable cytosol, PAR-2 predominantly associated with the plasma albeit weakly. Immunostaining for S-100 served as a positive membrane and a population of small juxta-nuclear vesicular control. Importantly, no appreciable staining was observed in structures that were reminiscent of endosomes/lysosomes. the absence of PAR antibodies or when chondrocytes were Together, these data demonstrate that PARs are both differ- treated with pre-immune serum at equivalent dilutions (data entially expressed and localized in human chondrocytes with Page 43

KIRILAK et al: EFFECTS OF FIBRIN SEALANT ON HUMAN CHONDROCYTES

PAR-1 being the major isotype expressed (PAR-1>PAR- during the course of this investigation, thrombin was shown 3>PAR-2>PAR-4). to directly stimulate the migration of osteogenic cultures of primary bone marrow cells to bony wound sites (25). Our 2+ Thrombin and PAR-1 agonists induce [Ca ]i influx in human in vitro MACI co-culture experiments demonstrate that the FS chondrocytes. The primary upstream signalling pathway of Tisseel induces a strong chemoattractive response in primary the PARs includes intracellular calcium mobilization (39). human chondrocytes. Significant migratory activity of 2+ Given that [Ca ]i mobilization have been well-documented to chondrocytes from the collagen scaffold towards the FS was correlate with cell growth and proliferation (20) we hypo- observed within a 24-h co-culture; however, there was little thesised that the thrombin-induced proliferation might reflect evidence of chondrocytic cell in-growth into the glue, changes in intracellular calcium signalling, possibly via inter- consistent with previous reports (10,47). While we did note actions with PARs. To explore this notion, we monitored for some marginal chondrocyte penetration into the FS at day 15, changes in free cytosolic calcium concentration in response this probably represented the onset of sealant disintegration to thrombin and specific PAR agonist peptides using the Ca2+ rather than a true reflection of cell in-growth. Clearly the indicator Fura-2. Thrombin at a concentration of 1 U/ml precise FS factor(s) conferring these migratory effects remain 2+ to be clarified and will be the focus of future research. None- elicited a large [Ca ]i elevation within a few seconds (mean amplitude of fluorescence ratio transient: 0.65±0.09; mean time theless, given thrombin is the active component of commercial to peak: 22.0±1.4 sec, n=8) in approximately 75% of human FS (Tisseel), together with its known chemotatic action; we chondrocytes examined, before gradually decaying (mean half speculate that this protease is the likely cause of the chondro- decay time: 32.4±4.6 sec; n=8) to near-baseline level (Fig. 6A). cytic migration observed in this study. In addition to promoting chemotaxis, thrombin has been Similarly, the PAR-1 agonist (SFLLRN-NH2, 400 μM) induced a significant and abrupt increase in intracellular [Ca2+] in an shown to stimulate the proliferation of a variety of cell types equivalent proportion of human chondrocytes. However, the (20,26,27,48-50). Consistently, our dose- and time-course- PAR-1-induced Ca2+ transients were significantly smaller in dependent studies chondrocyte conclusively demonstrate that amplitude (mean amplitude of fluorescence ratio transient; thrombin directly induces chondrocyte proliferation. Cellular 0.11±0.01; p=0.0002, n=6) than that elicited by thrombin, but proliferation was observed 24-h post-stimulation with a maximum response at 10 U/ml thrombin, suggesting a exhibited a steeper [Ca2+] elevation phase (mean time to peak: threshold response concentration close to this point. Moreover, 8.5±1.6 sec, p<0.0001, n=6). Nonetheless, the decay phase was we demonstrate that this increase in cellular proliferation comparable to the thrombin-induced Ca2+ transients (mean correlated with the mobilization of intracellular calcium. half decay time: 28.5±5.1 sec, p=0.58, n=6) (Fig. 6B). Similar responses were also observed using the TFLLRN-NH PAR-1 It is well-established that changes in cellular calcium elicit 2 several physiological sequelae, one of which is to stimulate cell agonist peptide (data not shown). In contrast to PAR-1 agonist mitotic activity. Thrombin has previously been shown to induce peptides, the PAR-4 (n=6) agonist peptide failed to elicit any the proliferation of various cell types via the modulation of visible Ca2+ response in human chondrocytes (Fig. 6B). intracellular calcium (28). In osteoblastic cells, this calcium Predictably, the PAR-3 agonist (n=5) did not evoke calcium influx is mediated largely through the interaction of thrombin mobilization, consistent with its reported inability to activate and activation of PARs suggesting that a similar mechanism PAR-3 and low-affinity to thrombin (40). Together, the may exist in ontogenically-related chondrocytes (20,27,33,35). analogies in [Ca2+] responses between thrombin and PAR-1 i Indeed, our expression studies document, for the first time, agonists hint that the thrombin-mediated effects on chondro- the expression of PARs in human chondrocytes thus extending cytes might act via PAR-1 signalling. chondrocytes to the list of PAR-expressing cells. PAR-1 was identified as the predominant isotype followed by PAR-3, Discussion PAR-2 and PAR-4. This finding is in accordance with the The diverse application of FSs has been advocated by expression patterns reported for other mesenchymal-lineage numerous studies from skin grafts through to attachment of cells (20,26,32,33). osteochondral fragments (2-4,41). We and others have Thrombin is known to activate PARs 1, 3 and 4 in various previously reported the use of FSs as an adjunct for ACI to cell types, whereas PAR-2 is activated by trypsin and tryptase secure scaffold fixation and promote chondrogenesis (20,33,51,52). Jenkins et al (53) demonstrated in osteoblasts (12,13,15,42-44). Moreover, our in vivo studies indicate that that intracellular calcium underwent a sustained rise after chondrocytes migrate towards the FS during osteochondral treatment with PAR-1-activating peptide, whereas thrombin repair suggesting that it may possess chemotactic potential. caused a sharp peak followed by a rapid return to baseline. 2+ The purpose of this study was to assess the bioactivity of the In contrast, our results show sustained [Ca ]i after thrombin commercial FS (Tisseel) on chondrocytes in vitro. Collectively, treatment, and a sharp transient increase after PAR-1- our data affirms that this FS not only promotes chondrocyte activating peptide treatment. It has been shown that ligand migration but also stimulates their proliferation. cross-reactivity exists in PARs whereby PAR-1-activating Several bioactive components constitute commercial and peptides are able to activate both PAR-1 and -2 (54). This is autologous FSs including fibrinogen, Factor XIII, fibronectin, contradictory to the data reported here. It is possible that the 2+ aprotinin and thrombin. Of these, thrombin has previously been elevated and sustained [Ca ]i response observed after shown to support the migration for a number of cell types thrombin treatment (in comparison to the PAR-1 agonist including monocytes (45), macrophages (24), neutrophils responses) is due to a number of factors. This discrepancy (38), fibroblasts (23) and endothelial cells (46). In addition, may be explained by either the presence of concurrent Page 44

INTERNATIONAL JOURNAL OF MOLECULAR MEDICINE 17: 551-558, 2006 intermolecular PAR activation between the tethered ligand 10. Brittberg M, Sjogren-Jansson E, Lindahl A and Peterson L: of PAR-1 and neighboring PARs (i.e. PAR-2) (55); combined Influence of fibrin sealant (Tisseel) on osteochondral defect repair in the rabbit knee. Biomaterials 18: 235-242, 1997. thrombin activation of PARs 1, 3, and 4; downstream interplay 11. Abiraman S, Varma HK, Umashankar PR and John A: Fibrin between activated G-protein-coupled PAR pathways; or the glue as an osteoinductive protein in a mouse model. Biomaterials possible presence of an unidentified chondrocyte thrombin- 23: 3023-3031, 2002. 12. Cherubino P, Grassi FA, Bulgheroni P and Ronga M: Autologous sensitive PAR. Based on previous studies on sustained chondrocyte implantation using a bilayer collagen membrane: a 2+ elevation of [Ca ]i after PAR-2 treatment (33), the former preliminary report. J Orthop Surg 11: 10-15, 2003. 13. Neovius EB and Kratz G: Tissue engineering by cocultivating two are considered the more likely possibilities. human elastic chondrocytes and keratinocytes. Tissue Eng 9: The apparent lack of calcium sensitivity observed upon 365-369, 2003. addition of the PAR-4 agonist peptide also implies that PAR-4 14. Visna P, Pasa L, Adler J, Folvarsky J and Horky D: Treatment of deep chondral defects of the knee using autologous chondro- may not be crucial to chondrocyte physiology. This notion is cytes cultured on a support - preparation of the cartilage graft. supported by the comparatively weak expression levels of Acta Chir Orthop Traumatol Cech 70: 350-355, 2003. PAR-4 detected in human chondrocytes. Moreover, these 15. Willers C, Chen J, Wood D, Xu J and Zheng MH: Autologous chondrocyte implantation with collagen bioscaffold for the findings are in accordance with calcium mobilization studies treatment of osteochondral defects in rabbits. Tissue Eng 11: conducted with PAR-4 agonists in gingival fibroblasts which 1065-1076, 2005. also exhibit strong PAR-1 and -3 but weak PAR-2 and -4 16. Powers JC and Kam CM: Synthetic Substrates and Inhibitors of Thrombin. Plenum Press, New York, pp117-158, 1992. expression (56). Further studies will be required to unravel 17. Tani K, Yasuoka S, Ogushi F, Asada, K, Fujisawa K, Ozaki T, the precise roles of specific PAR isoforms in chondrocytes. Sano N and Ogura T: Thrombin enhances lung fibroblast In summary, we demonstrate that the commercial FS proliferation in bleomycin-induced pulmonary fibrosis. Am J Respir Cell Mol Biol 5: 34-40, 1991. Tisseel promotes the migration and proliferation of primary 18. Borrelli V, Sterpetti AV, Coluccia P, Randone B, Cavallaro A, human chondrocytes. Moreover, we demonstrate that the Santoro D'Angelo L and Cucina A: Bimodal concentration- thrombin component alone is sufficient to stimulate chondro- dependent effect of thrombin on endothelial cell proliferation and growth factor release in culture. J Surg Res 100: 154-160, cyte proliferation and elicit intracellular calcium mobilization. 2001. In addition, we document for the first time, the expression and 19. Kaufmann K and Thiel G: Epidermal growth factor and localisation of PARs in chondrocytes and provide evidence to thrombin induced proliferation of immortalized human keratinocytes is coupled to the synthesis of Egr-1, a zinc finger suggest that PAR-1 is the primary thrombin-acting PAR in transcriptional regulator. J Cell Biochem 85: 381-391, 2002. chondrocytes and, thus, might account for the observed 20. Abraham, LA and MacKie EJ: Modulation of osteoblast-like migratory and proliferative responses. This hypothesis will cell behavior by activation of protease-activated receptor-1. J Bone Miner Res 14: 1320-1329, 1999. form the basis of more detailed studies in the future. 21. Song SJ, Pagel CN, Campbell TM, Pike RN and Mackie EJ: Nonetheless, the data presented in this study endorse the use of The role of protease-activated receptor-1 in bone healing. Am J FSs in autologous chondrocyte implantation for cartilage Pathol 166: 857-868, 2005. 22. Bar-Shavit R, Kahn A, Fenton JW and Wilner GD: Chemotactic injury. response of monocytes to thrombin. J Cell Biol 96: 282-285, 1983. 23. Brown LF, Lanir N, McDonagh J, Tognazzi K, Dvorak AM and Acknowledgements Dvorak HF: Fibroblast migration in fibrin gel matrices. Am J Pathol 142: 273-283, 1993. This study was supported by grants from the National Health 24. Ciano PS, Colvin RB, Dvorak AM, McDonagh J and Dvorak HF: Macrophage migration in fibrin gel matrices. Lab Invest 54: and Medical Research Council (NHMRC) of Australia. 62-70, 1986. 25. Karp JM, Tanaka TS, Zohar R, Sodek J, Shoichet, MS, Davies JE References and Stanford WL: Thrombin mediated migration of osteogenic cells. Bone 37: 337-348, 2005. 26. Chinni C, De Niese MR, Tew DJ, Jenkins AL, Bottomley SP 1. Jackson MR: Fibrin sealants in surgical practice: an overview. and Mackie EJ: Thrombin, a survival factor for cultured Am J Surg 182: 1S-7S, 2001. myoblasts. J Biol Chem 274: 9169-9174, 1999. 2. Currie LJ, Sharpe JR and Martin R: The use of fibrin glue in 27. Pagel CN, De Niese MR, Abraham LA, Chinni C, Song SJ, skin grafts and tissue-engineered skin replacements: a review. Pike RN and Mackie EJ: Inhibition of osteoblast apoptosis by Plast Reconstr Surg 108: 1713-1726, 2001. thrombin. Bone 33: 733-743, 2003. 3. Albala DM: Fibrin sealants in clinical practice. Cardiovasc Surg 28. Mackie EJ, Pagel CN, Smith R, De Niese MR, Song SJ and 11 (Suppl 1): 5-11, 2003. Pike RN: Protease-activated receptors: a means of converting 4. Canonico S: The use of human fibrin glue in the surgical extracellular proteolysis into intracellular signals. IUBMB Life operations. Acta Biomed Ateneo Parmense 74 (Suppl 2): 21-25, 53: 277-281, 2002. 2003. 29. Ishihara H, Connolly AJ, Zeng D, Kahn ML, Zheng YW, 5. Kaplonyi G, Zimmerman I, Frenyo AD, Farkas T and Nemes G: Timmons C, Tram T and Coughlin SR: Protease-activated The use of fibrin adhesive in the repair of chondral and receptor 3 is a second thrombin receptor in humans. Nature 386: osteochondral injuries. Injury 19: 267-272, 1988. 502-506, 1997. 6. Ono K, Shikata J, Shimizu K and Yamamuro T: Bone-fibrin 30. Vu TK, Hung DT, Wheaton VI and Coughlin SR: Molecular mixture in spinal surgery. Clin Orthop Relat Res 275: 133-139, cloning of a functional thrombin receptor reveals a novel 1992. proteolytic mechanism of receptor activation. Cell 64: 1057-1068, 7. Homminga GN, Bulstra SK, van der Linden AJ and 1991. Bouwmeester PS: Perichondral grafting for cartilage lesions of 31. Akers IA, Parsons M, Hill MR, Hollenberg MD, Sanjar S, the knee. J Bone Joint Surg Br 72: 1003-1007, 1990. Laurent GJ and McAnulty RJ: Mast cell tryptase stimulates 8. Homminga GN, Buma P, Koot HW, van der Kraan PM and human lung fibroblast proliferation via protease-activated van den Berg WB: Chondrocyte behavior in fibrin glue in vitro. receptor-2. Am J Physiol Lung Cell Mol Physiol 278: L193-L201, Acta Orthop Scand 64: 441-445, 1993. 2000. 9. Hendrickson DA, Nixon A J, Erb HN and Lust G: Phenotype 32. Chinni C, De Niese MR, Jenkins AL, Pike RN, Bottomley SP and biological activity of neonatal equine chondrocytes cultured and Mackie EJ: Protease-activated receptor-2 mediates in a three-dimensional fibrin matrix. Am J Vet Res 55: 410-414, proliferative responses in skeletal myoblasts. J Cell Sci 113: 1994. 4427-4433, 2000. Page 45

KIRILAK et al: EFFECTS OF FIBRIN SEALANT ON HUMAN CHONDROCYTES

33. Abraham LA, Chinni C, Jenkins AL, Lourbakos A, Ally N, 46. Nehls V and Herrmann R: The configuration of fibrin clots Pike RN and Mackie EJ: Expression of protease-activated determines capillary morphogenesis and endothelial cell receptor-2 by osteoblasts. Bone 26: 7-14, 2000. migration. Microvasc Res 51: 347-364, 1996. 34. Brass LF and Molino M: Protease-activated G protein-coupled 47. Gille J, Meisner U, Ehlers EM, Muller A, Russlies M and receptors on human platelets and endothelial cells. Thromb Behrens P: Migration pattern, morphology and viability of cells Haemost 78: 234-241, 1997. suspended in or sealed with fibrin glue: a histomorphologic 35. Asokananthan N, Graham PT, Fink J, Knight DA, Bakker AJ, study. Tissue Cell 37: 339-348, 2005. McWilliam AS, Thompson PJ and Stewart GA: Activation of 48. Frost A, Jonsson KB, Ridefelt P, Nilsson O, Ljunghall S and protease-activated receptor (PAR)-1, PAR-2, and PAR-4 Ljunggren O: Thrombin, but not bradykinin, stimulates stimulates IL-6, IL-8, and prostaglandin E2 release from human proliferation in isolated human osteoblasts, via a mechanism not respiratory epithelial cells. J Immunol 168: 3577-3585, 2002. dependent on endogenous prostaglandin formation. Acta Orthop 36. Pavlos NJ, Xu J, Riedel D, Yeoh JS, Teitelbaum SL, Jahn R, Scand 70: 497-503, 1999. Papadimitriou JM, Ross FP and Zheng MH: Rab3D regulates a 49. De Niese MR, Chinni C, Pike RN, Bottomley SP and Mackie EJ: novel vesicular trafficking pathway that is required for Dissection of protease-activated receptor-1-dependent and osteoclastic bone resorption. Mol Cell Biol 25: 5253-5269, -dependent responses to thrombin in skeletal myoblasts. Exp 2005. Cell Res 274: 149-156, 2002. 37. Wang HS, Li F, Runge MS and Chaikof EL: Endothelial cells 50. Gruber R, Jindra C, Kandler B, Watzak G, Fischer MB and exhibit differential chemokinetic and mitogenic responsiveness Watzek G: Proliferation of dental pulp fibroblasts in response to to alpha-thrombin. J Surg Res 68: 5253-5264, 1997. thrombin involves mitogen-activated protein kinase signalling. 38. Jenkins AL, Howells GL, Scott E, Le Bonniec BF, Curtis MA Int Endod J 37: 145-150, 2004. and Stone SR: The response to thrombin of human neutrophils: 51. Corvera CU, Dery O, McConalogue K, Gamp P, Thoma M, evidence for two novel receptors. J Cell Sci 108: 3059-3066, Al-Ani B, Caughey GH, Hollenberg MD and Bunnett NW: 1995. Thrombin and mast cell tryptase regulate guinea-pig myenteric 39. Dery O, Corvera CU, Steinhoff M and Bunnett NW: Proteinase- neurons through proteinase-activated receptors-1 and -2. J activated receptors: novel mechanisms of signaling by serine Physiol 517: 741-756, 1999. proteases. Am J Physiol 274: C1429-C1452, 1998. 52. Molino M, Barnathan ES, Numerof R, Clark J, Dreyer M, 40. Owen WG: PAR-3 is a low-affinity substrate, high affinity Cumashi A, Hoxie JA, Schechter N, Woolkalis M and Brass LF: effector of thrombin. Biochem Biophys Res Commun 305: Interactions of mast cell tryptase with thrombin receptors and 166-168, 2003. PAR-2. J Biol Chem 272: 4043-4049, 1997. 41. Shah MA, Ebert AM and Sanders WE: Fibrin glue fixation of a 53. Jenkins AL, Bootman MD, Berridge MJ and Stone SR: digital osteochondral fracture: case report and review of the Differences in intracellular calcium signaling after activation of literature. J Hand Surg [Am] 27: 464-469, 2002. the thrombin receptor by thrombin and agonist peptide in 42. Hutmacher DW, Goh JC and Teoh SH: An introduction to osteoblast-like cells. J Biol Chem 269: 17104-17110, 1994. biodegradable materials for tissue engineering applications. Ann 54. Blackhart BD, Emilsson K, Nguyen D, Teng W, Martelli AJ, Acad Med Singapore 30: 183-191, 2001. Nystedt S, Sundelin J and Scarborough RM: Ligand cross- 43. Kaplonyi G, Zimmermann I, Farkas T and Viola T: Repair of reactivity within the protease-activated receptor family. J Biol cartilage injuries and osteochondral fractures with fibrin glue. Chem 271: 16466-16471, 1996. Orv Hetil 125: 2237-2243, 1984. 55. O'Brien PJ, Molino M, Kahn M and Brass LF: Protease-activated 44. Visna P, Pasa L, Hart R, Kocis J, Cizmar I and Adler J: Treatment receptors: theme and variations. Oncogene 20: 1570-1581, of deep chondral defects of the knee using autologous chondro- 2001. cytes cultured on a support - results after one year. Acta Chir 56. Tanaka N, Morita T, Nezu A, Tanimura A, Mizoguchi I and Orthop Traumatol Cech 70: 356-362, 2003. Tojyo Y: Thrombin-induced Ca2+ mobilization in human gingival 45. Bar-Shavit Z, Teitelbaum SL, Reitsma P, Hall A, Pegg LE, Trial J fibroblasts is mediated by protease-activated receptor-1 (PAR-1). and Kahn AJ: Induction of monocytic differentiation and bone Life Sci 73: 301-310, 2003. resorption by 1,25-dihydroxyvitamin D3. Proc Natl Acad Sci USA 80: 5907-5911, 1983.

Chapter 4

Sheep MACI: Repair Assessment by Confocal Arthroscopy

Thesis publication #3: Jones CW, Willers C, et al. Matrix-induced Autologous Chondrocyte Implantation (MACI®) in Sheep: Objective Assessments Including Confocal Arthroscopy. Journal of Orthopaedic Research, 2007.

Page 46

STATEMENTS OF AUTHORSHIP CONTRIBUTION FROM CO-AUTHORS

Matrix-induced Autologous Chondrocyte Implantation (MACI®) in Sheep: Objective Assessments Including Confocal Arthroscopy.

Jones CW, Willers C, Keogh A, Smolinski D, Fick D, Yates P, Kirk TB, Zheng MH.

Published in the Journal of Orthopaedic Research 2007.

Jones CW (Research Collaborator) Major contribution to the planning, execution, analysis, and interpretation of the research. Major contribution to manuscript writing.

Signature of Principal Author: …………………………… Date: ……………………

Willers C (PhD Candidate) Moderate contribution to planning of research. Major contribution to execution (graft cultivation, supervised surgery, histology/immunohistochemistry) and analysis (scored LSCA and histology outcomes) of research. Moderate contribution to interpretation of research. Moderate contribution to manuscript writing.

Signature of Co-Author: ……………………..……………… Date: …………………

Keogh A (Research Collaborator) Major contribution to the execution of research (chief surgeon and LSCA operator). Minor contribution to manuscript writing

Signature of Co-Author: ……………………..……………… Date: …………………

Page 47

Smolinski D (Research Collaborator) Moderate contribution to planning of research (design and development of LSCA). Minor contribution to execution of research (assisted surgery and collected data). Minor contribution to manuscript writing.

Signature of Co-Author: ……………………..……………… Date: …………………

Fick D (Research Collaborator) Minor contribution to the execution of research (few MACI surgeries). Moderate contribution to interpretation of research (MRI data). Minor contribution to manuscript writing.

Signature of Co-Author: ……………………..……………… Date: …………………

Yates P (Research Collaborator) Minor contribution to the execution of research (few MACI surgeries). Minor contribution to manuscript writing.

Signature of Co-Author: ……………………..……………… Date: …………………

Kirk TB (Research Collaborator) Major contribution to the planning of research. Minor contribution to manuscript writing.

Signature of Co-Author: ……………………..……………… Date: …………………

Zheng MH (Supervisor) Major contribution to the planning of research. Minor contribution to manuscript writing, and acted as corresponding Author.

Signature of Co-Author: ……………………..……………… Date: ………………… Page 48

STATEMENTS OF PERMISSION FROM CO-AUTHORS

I, Chris Jones give permission to Craig Willers, second author of the paper: Matrix- induced Autologous Chondrocyte Implantation (MACI) in Sheep: Objective Assessments Including Confocal Arthroscopy, published in the Journal of Orthopaedic Research (2007), to include this paper as Chapter 4 of his PhD thesis entitled Matrix-induced Autologous Chondrocyte Implantation for Articular Cartilage Injury: Biology, Histology, and Clinical Outcomes.

Signature: …………………………………. Date: ………………………………….

I, Angus Keogh give permission to Craig Willers, second author of the paper: Matrix- induced Autologous Chondrocyte Implantation (MACI) in Sheep: Objective Assessments Including Confocal Arthroscopy, published in the Journal of Orthopaedic Research (2007), to include this paper as Chapter 4 of his PhD thesis entitled Matrix-induced Autologous Chondrocyte Implantation for Articular Cartilage Injury: Biology, Histology, and Clinical Outcomes.

Signature: …………………………………. Date: …………………………………

I, Daniel Smolinski give permission to Craig Willers, second author of the paper: Matrix-induced Autologous Chondrocyte Implantation (MACI) in Sheep: Objective Assessments Including Confocal Arthroscopy, published in the Journal of Orthopaedic Research (2007), to include this paper as Chapter 4 of his PhD thesis entitled Matrix-induced Autologous Chondrocyte Implantation for Articular Cartilage Injury: Biology, Histology, and Clinical Outcomes.

Signature: …………………………………. Date: …………………………………

Page 49

I, Daniel Fick give permission to Craig Willers, second author of the paper: Matrix- induced Autologous Chondrocyte Implantation (MACI) in Sheep: Objective Assessments Including Confocal Arthroscopy, published in the Journal of Orthopaedic Research (2007), to include this paper as Chapter 4 of his PhD thesis entitled Matrix-induced Autologous Chondrocyte Implantation for Articular Cartilage Injury: Biology, Histology, and Clinical Outcomes.

Signature: …………………………………. Date: .………………………………

I, Piers Yates give permission to Craig Willers, second author of the paper: Matrix- induced Autologous Chondrocyte Implantation (MACI) in Sheep: Objective Assessments Including Confocal Arthroscopy, published in the Journal of Orthopaedic Research (2007), to include this paper as Chapter 4 of his PhD thesis entitled Matrix-induced Autologous Chondrocyte Implantation for Articular Cartilage Injury: Biology, Histology, and Clinical Outcomes.

Signature: …………………………………. Date: …………………………………

I, Brett Kirk give permission to Craig Willers, second author of the paper: Matrix- induced Autologous Chondrocyte Implantation (MACI) in Sheep: Objective Assessments Including Confocal Arthroscopy, published in the Journal of Orthopaedic Research (2007), to include this paper as Chapter 4 of his PhD thesis entitled Matrix-induced Autologous Chondrocyte Implantation for Articular Cartilage Injury: Biology, Histology, and Clinical Outcomes.

Signature: …………………………………. Date: …………………………………

Page 50

I, Ming-Hao Zheng give permission to Craig Willers, second author of the paper: Matrix-induced Autologous Chondrocyte Implantation (MACI) in Sheep: Objective Assessments Including Confocal Arthroscopy, published in the Journal of Orthopaedic Research (2007), to include this paper as Chapter 4 of his PhD thesis entitled Matrix-induced Autologous Chondrocyte Implantation for Articular Cartilage Injury: Biology, Histology, and Clinical Outcomes.

Signature:…………………………………. Date:……………………………………

Page 51

Matrix-Induced Autologous Chondrocyte Implantation in Sheep: Objective Assessments Including Confocal Arthroscopy

C. W. Jones,1,2 C. Willers,2 A. Keogh,2 D. Smolinski,1 D. Fick,2 P. J. Yates,2 T. B. Kirk,1 M. H. Zheng2 1School of Mechanical Engineering, University of Western Australia, 35 Stirling Highway, Crawley WA, 6009, Australia 2School of Pathology and Surgery, Department of Orthopaedics, University of Western Australia, 2nd Floor M-block QEII Medical Centre, Nedlands, Perth WA, 6009, Australia

Received 18 August 2006; accepted 10 July 2007 Published online in Wiley InterScience (www.interscience.wiley.com). DOI 10.1002/jor.20502

ABSTRACT: The assessment of cartilage repair has largely been limited to macroscopic observation, magnetic resonance imaging (MRI), or destructive biopsy. The aims of this study were to establish an ovine model of articular cartilage injury repair and to examine the efﬁcacy of nondestructive techniques for assessing cartilage regeneration by matrix-induced autologous chondrocyte implantation (MACI). The development of nondestructive assessment techniques facilitates the monitoring of repair treatments in both experimental animal models and human clinical subjects. Defects (Ø 6 mm) were created on the trochlea and medial femoral condyle of 21 sheep randomized into untreated controls or one of two treatment arms: MACI or collagen-only membrane. Each group was divided into 8-, 10-, and 12-week time points. Repair outcomes were examined using laser scanning confocal arthroscopy (LSCA), MRI, histology, macroscopic ICRS grading, and biomechanical compression analysis. Interobserver analysis of the randomized blinded scoring of LSCA images validated our scoring protocol. Pearson correlation analysis demonstrated the correlation between LSCA, MRI, and ICRS grading. Testing of overall treatment effect independent of time point revealed signiﬁcant differences between MACI and control groups for all sites and assessment modalities (Asym Sig < 0.05), except condyle histology. Biomechanical analysis suggests that while MACI tissue may resemble native tissue histologically in the early stages of remodeling, the biomechanical properties remain inferior at least in the short term. This study demonstrates the potential of a multisite sheep model of articular cartilage defect repair and its assessment via nondestructive methods. ß 2007 Orthopaedic Research Society. Published by Wiley Periodicals, Inc. J Orthop Res Keywords: cartilage repair; matrix-induced autologous chondrocyte impantation (MACI); confocal microscopy

INTRODUCTION ciated donor site morbidity on the tibia.2,3 While excellent clinical results have been achieved, It has long been recognized that once damaged, significant controversy remains regarding the articular cartilage may progress from difficult-to- histological nature of the tissue formed.4,5 treat lesions to osteoarthritis. The problems In conjunction with the International Cartilage associated in treating the initial lesions have Repair Society (ICRS), Brittberg and Winalski prompted the development of a number of appro- have standardized the clinical assessment, arthro- aches, most recently matrix-induced autologous scopic evaluation, and magnetic resonance imaging chondrocyte implantation (MACI).1,2 MACI uti- (MRI) of focal cartilage lesions and their repair.6 In lizes a type I/III collagen bioscaffold to traffic combination with clinical measures, the ICRS cultured autologous chondrocytes into the defect, system provides an overall assessment of cartilage thereby obviating periosteal harvesting and asso- repair, but provides no histological data. Histolog- ical assessment of cartilage repair tissue distin- 4,7 This article includes Supplementary Material available via guishes between tissue types. Previous studies the Internet at http://www.interscience.wiley.com/jpages/0736- utilized mechanical biopsy to provide tissue sam- 0266/suppmat. ples for histological analysis.3,4,8,9 A limited Correspondence to: M.H. Zheng (Telephone: þ61 8 93464050; amount of data exist regarding the cellular and Fax: þ61 8 93463210; E-mail: [email protected]) ß 2007 Orthopaedic Research Society. Published by Wiley Periodicals, microstructural nature of MACI repair tissue due Inc. to early demonstrations of clinical efficacy and

JOURNAL OF ORTHOPAEDIC RESEARCH 2007 Page 52

JONES ET AL. reluctance to take biopsy samples in human genrath, Germany). Six weeks after surgery, implanta- patients. Indeed, Poole cautioned against mechan- tion was undertaken via the previous parapatellar ical biopsy of articular grafts due to the danger of approach with repunching and debridement performed sustaining further damage.5 Ergo, the develop- to maintain a well-circumscribed defect and remove any ment of the laser scanning confocal arthroscope fibrous repair tissue. The chondrocyte-seeded collagen membrane was shaped to match the defect geometry, and (LSCA), an optical biopsy tool capable of non- 1 press-fit onto the floor of the defect with Tisseel fibrin destructively imaging articular cartilage repair 10,11 sealant (Baxter, Vienna, Austria). The second treatment tissue in situ and in vivo. The need for arm of the study involved implantation of defects with controlled trials of alternative repair techniques acellular type I/III collagen membrane in an identical indicates the further development of reliable large manner to MACI surgical protocol. Defect-only controls animal models. were established as untreated partial-thickness lesions In this study, our objectives were to establish a in comparable sites. Euthanasia of sheep was performed large animal model of articular cartilage injury at each specified time point and imaging carried out repair and to analyze the effectiveness of MACI immediately. objectively using the nondestructive techniques of LSCA and MRI in comparison to the gold standard ICRS Cartilage Repair Assessment of histological assessment. Mechanical analysis of ICRS grading was conducted during open joint exami- repair tissue was also conducted. nation (prior to LSCA imaging) by a single orthopedic surgeon. Grading was corroborated by the consultant MATERIALS AND METHODS histopathologists on en bloc specimens. Cartilage repair tissue was evaluated according to the ICRS Cartilage Experimental Design Injury Evaluation Package (Protocol A) considering Twenty-one sheep were randomly divided into three macroscopic appearance, volume of defect filled, and groups. Defect-only control animals were compared to integration with adjacent cartilage.6 animals surgically treated with MACI or type I/III collagen membrane alone. After surgical intervention, Laser Scanning Confocal Arthroscopy animals were divided into three time periods for final assessment at 8 (2 MACI, 2 collagen, 2 control), 10 LSCA immediately followed ICRS assessment and (3 MACI, 3 collagen, 3 control), and 12 weeks enabled imaging in situ without mechanical biopsy or (2 MACI, 2 collagen, 2 control; see also Supple- tissue processing. The LSCA uses a proprietary optical mentary Material). fiber scanner both performing the laser delivery (488 nm argon-ion laser) and acting as the confocal pinhole and Surgical Procedure filtration mechanism.10,11,13 The LSCA is a 4.4-mm diameter arthroscopically mounted miniaturized con- All operations were conducted under strict guidelines of focal microscope providing an xy spatial resolution of 2 mm the National Health and Medical Research Council across a field of view of approximately 500 500 mm (Canberra, Australia). All surgery was performed by the (512 512 pixels) and focal plane penetration of 200 mmat orthopedic surgeons named as coauthors. Animals were 2 Hz (see Supplementary Material). The fluorophores sedated and anaesthetized, administered with antibiotic were acridine orange (0.5 g/L, 30 mL, 30–40 min) and prophylaxis, and provided with pre- and postoperative fluorescein (5 g/L, 30mL, 40–50 min; Molecular Probes analgesia (ketamine 11 mg/kg, xylaxine 0.22 mg/kg, Inc., Eugene, OR). Stains were prepared immediately keflin 1 mg, carprofin 2–4 mg/kg, buprenorphine/temgesic prior to imaging in 0.9% phosphate-buffered saline (PBS) 1 ml IM). Unilateral stifle joints were surgically accessed at physiological pH and temperature (378C) and stored via a medial parapatellar approach, and standardized away from light. Optimal staining concentration, vol- partial thickness (1.5 mm) trochlea and medial femoral ume, and times were established during a previous condyle defects were created using a custom-designed study.10 Lavage with 0.9% PBS (300 mL) was performed 6 mm chondral punch. Nonweightbearing cartilage from prior to imaging until all excess stain was removed. the lateral supracondyle was harvested for cell culture. More than 2,500 individual confocal images were edited Defect debris was removed and the base of the defect to three representative images for each site, blinded, leveled via surgical curette. Following defect creation computer randomized, and distributed to two histopa- the wound was closed via interrupted capsular (1 vicryl) thologists and three orthopedic surgeons for modality and subcuticular sutures (3-0 monocryl). scoring (Table 1). Blinding of the treatment group Animals were not immobilized, but movement was and time point was maintained by code known only to confined for 24 h postoperatively. Using a previously the chief investigator for the duration of the analysis. described protocol for cultivation of rabbit and human chondrocytes,3,12 autologous chondrocytes were isolated 6 Histology and Immunohistochemistry under high sterility conditions, expanded to >5 10 cells, and seeded onto collagen membrane (porcine- Distal femora were fixed then decalcified with 10% derived type I/III collagen membrane; Matricel, Herzo- formic acid. Specimen blocks (8 8 mm) were prepared

JOURNAL OF ORTHOPAEDIC RESEARCH 2007 DOI 10.1002/jor Page 53

MACI IN SHEEP

Table 1. Modality Scoring and Overall Repair Assessment Grading

Modality Scoring

MRI Cartilage Repair Assessment LSCA Cartilage Repair Assessment

Criteria Score Criteria Score 1. Infill of the repair site 1. Cellularity Excellent: complete 4 High density 4 Good: >50% of the defect 3 Moderate density 3 Fair: <50% 2 Low density 2 Poor: a full-thickness defect 1 Acellular 1 2. Surface continuity 2. Spatial uniformity of cells Excellent: smooth 4 Uniform 4 Good: focal changes only 3 Almost uniform 3 Fair: <50% fibrillation 2 Heterologous uniformity 2 Poor: ulceration or delamination 1 Non-uniform 1 3. Signal at the repair vs. adjacent AC 3. Cell distribution Excellent: >75% intensity 4 Diffuse with paired 3 Good: 75% intensity 3 Diffuse without paired 2 Fair: 50% intensity 2 Predominant clustering 1 Poor: 25% intensity 1 4. Structure 4. Cell morphology Excellent: >75% homogeneous 4 Predominantly spherical 3 Good: >50% homogeneous 3 Mixed morphology 2 Fair: heterogeneous with no clefts 2 Predominantly spindle-shaped 1 Poor: heterogeneous with clefts 1 5. Border integration 5. Visibility of Collagen Fibers Excellent: complete integration 4 No fibers 3 Good: minor gap 3 Some fibers 2 Fair: incomplete, gap visible 2 Highly fibrous 1 Poor: incomplete, visible defect 1 6. Subchondral lamina Excellent: fully reconstituted 4 Good: >50% intact 3 Fair: <25% intact 2 Poor: no visible lamina 1 7. Subchondral bone condition Excellent: normal, intact, no edema 4 Good: edema <1 cm from lamina 3 Fair: >1 cm from lamina 2 Poor: cysts, sclerosis 1 8. Joint effusion 9. Synovitis None 4 None 4 Mild 3 Mild 3 Moderate 2 Moderate 2 Severe 1 Severe 1

ORA Grading LSCA MRI WAZ ICRS

Grade I: normal 16 34 20 12 Grade II: nearly normal 15–12 25–33 12–19 8–11 Grade III: abnormal 8–11 16 –24 7–11 4–7 Grade IV: severely abnormal <8 <16 <7 <3 for processing by osteotomizing the defect center. (H&E), alcian blue (AB), then type II collagen immuno- Samples were dehydrated by a graded series of alcohol histochemistry (with IgG control) was performed as and xylene washes and parafﬁn-embedded. Sections previously described.3 Sections were further stained, were cut to 5 mm and stained with hematoxylin and eosin viewed under light microscope, and scored using the

DOI 10.1002/jor JOURNAL OF ORTHOPAEDIC RESEARCH 2007 Page 54

JONES ET AL.

WAZ semiquantitative scoring system as previously Table 2. Pearson Correlations for Condyle Treatment described.3 The score is based on percentage defect Groups (ICRS vs. WAZ vs. LSCA vs. MRI), Demonstrating filling, articular surface continuity, tissue integration, Significant Correlation Between LSCA and MRI and cellular morphology of cartilage regeneration, and Between ICRS and MRI (p < 0.05) matrix staining (collagen II, proteoglycan). Condyle MRI LSCA WAZ Magnetic Resonance Imaging ICRS r2 0.608** 0.212 0.349 MRI was conducted using a 1.5 T closed unit without p-value 0.007 0.369 0.132 extremity coil (Siemens Vision; Siemens, Erlangen, WAZ r2 0.294 0.007 Germany). Proton density (PD) imaging was applied to p-value 0.236 0.977 make the native joint cartilage bright, thus providing LSCA r2 0.469* good differential contrast against the black subchondral p-value 0.050 bone and underlying marrow. The MRI techniques used were in line with the recommendations of the Articular Trochlea LSCA WAZ Cartilage Imaging Group of the ICRS.14 Sequences included coronal T2 fat-suppressed imaging (TR/TE ICRS r2 0.589** 0.367 4650/81, FOV 14 cm, slice thickness 3.0 mm); coronal p-value 0.006 0.112 PD imaging (TR/TE 2060/34, FOV 14 cm, slice thickness WAZ r2 0.341 3.0 mm); sagittal PD imaging (TR/TE 2720/32, FOV p-value 0.141 14 cm, slice thickness 4.0 mm); sagittal T2 fat-suppressed imaging (TR/TE 3400/72, FOV 14 cm, slice Pearson Correlations for Trochlea Treatment Groups (ICRS vs. WAZ vs. LSCA), Demonstrating Significant Correlation thickness 4.0 mm); and axial PD fat-suppressed imaging Between LSCA and ICRS (p < 0.05). (TR/TE 3000/38, FOV 15 cm, slice thickness 3.0 mm). A *Correlation is significant at the 0.05 level (two-tailed). targeted approach was used whereby high-resolution **Correlation is significant at the 0.01 level (two-tailed). images in two planes were performed tangential to the graft site. Similar to the MOCART protocol for assessing cartilage repair, an experienced musculoskeletal radi- MT1-Z6) actuated by a servo-motorized micrometer ology consultant scored the MR scans for infill, surface (Thorlabs Z612B DC–90N). Test position recording continuity, signal, structure, border integration, sub- was performed via high precision (0.05 mm) rotary chondral lamina, subchondral bone condition, joint optical encoder. 15 effusion, and synovitis (Table 1). Samples were placed in the centre of ; 6mm impermeable load aluminium platens with CNC mac- Overall Repair Assessment hined guides to ensure platen surfaces remained parallel. Under partially confined compression, the top and Overall repair assessment (ORA) was made utilizing an bottom surfaces were fixed by sand paper to the platens adaptation of the ICRS grade I to IV classification while the sides were free to expand laterally. Sufficient system to compare assessment modalities for each acceleration and deceleration tails ensured loading at a treatment group and time point. Histology, ICRS macro- constant nominal strain rate. Samples were axially scopic grading, LSCA, and MRI modality scores were compressed by 20% of their initial height and the cycle converted into an ORA grade I to IV (Table 1) for repeated until a steady state was reached (about 5 Pearson correlation analysis (Table 2). cycles), with the subsequent compression loading cycle Overall Treatment Effect recorded. Backlash (<8.0 mm) was avoided by reposition- ing the stage prior to each cycle. Results are presented as Analysis of the overall treatment effect was conducted average vertical Lagrange stress (sz) versus relative by comparing mean modality scores for each treatment displacement (L) because the partially confined config- group both dependent and independent of time point. uration assumes tissue anisotropy. L is therefore ana- logous (but not equal) to the vertical extension ratio lz Mechanical Testing calculated in unconfined compression. For 2H (the pretest sample height) and 2h (the compressed height), Cylindrical test specimens were extracted using a L is defined as h/H. s is defined as vertical reaction force 1 z ; 2 mm trephine biopsy punch (Stiefel Laboratories, (Fz) divided by cross-sectional area (Az). The stiffness Inc.) such that cylinder axes were aligned perpendicular relationship is thus defined as: to the chondral surface. Final test specimens were s F H produced by trimming remaining subchondral bone z ¼ z under microscopic guidance; specimen height was then L Axh recorded. Samples were fast frozen and stored at 40C8, then thawed by immersion in 0.9% saline at room Statistical Methods temperature for 30 min prior to testing. The compression rig utilized a load cell (Entran1 ELFM-B1-10L- Interobserver variability from the randomized, blinded 50N) atop a linear translation stage (Thorlabs, Inc. LSCA scoring data was examined using an intraclass

JOURNAL OF ORTHOPAEDIC RESEARCH 2007 DOI 10.1002/jor Page 55

MACI IN SHEEP correlation analysis (ICC) performed using the statis- (p < 0.05; Fig. 2C). For condyle repairs, no signifi- tical analysis package SPSS (SPSS 11.0, Chicago, IL). cant difference was seen between the MACI and An a, fully crossed two-way ANOVA mixed model ICC collagen-only scaffold treatment groups in com- design was employed, generating a one-way random parison to the defect-only controls (Fig. 2D). effect model of intraclass correlation expressed as ICC Interobserver variability ICC analysis performed with 95% confidence interval (CI). The ICC is a on the randomized and blinded LSCA scores reliability procedure used to estimate interrater reliability based on mean squares obtained by applying demonstrated substantial agreement between sequential analysis of variance (perfect agreement, 1.0; observers (ICC ¼ 0.70; 95%CI, 0.63–0.76). The almost perfect agreement, 1.0–0.81; substantial agree- average measure of intraclass correlation across ment, 0.8–0.61; fair agreement, 0.6–0.41; moderate all observers suggested almost perfect agreement agreement, 0.4–0.21; poor agreement, 0.2–0.01; chance (ICC ¼ 0.92; 95% CI, 0.89–0.94). LSCA confocal agreement, 0.0; and perfect disagreement, 1.0).16 For images revealed clear differences between native each assessment modality, multiple pairwise Student t- cartilage, regenerated tissue, fibrocartilage repair, tests were used to assess differences between modality and degenerative tissue (Fig. 3). scores of treatment groups compared to control at each time point. Assessment of the overall treatment Histology effect compared mean modality scores for each treatment group (dependent and independent of time point) MACI-treated lesions were characterized by a using the Kruskal–Wallis test with a Bonferroni post mixed repair of hyalinelike cartilage and fibrocar- hoc test. Correlation between modalities was facilitated tilage with good infill and integration, pericellular by ORA Grades assigned to according to the protocol specific proteoglycan staining, and weak type II (Table 1). Pearson coefficients examined direct correla- collagen staining (Fig. 1C). Collagen-only bioscaf- tions between assessment modalities relative to site, fold or untreated control defect repairs showed treatment group, and time point for ORA grades I to IV. fibrous tissue repair with partial infill, poor RESULTS integration, poor architectural restoration, and no proteoglycan or type II collagen production. Examples of all imaging modalities are presented Cellular migration into the acellular collagen-only in the Supplementary Material accompanying this bioscaffold treatment group was observed across publication. all time points. Trochlea histology scores demonstrated improved tissue regeneration after MACI ICRS Cartilage Repair Assessment treatment from 8 to 12 weeks. The regenerated tissue in MACI-treated trochlear lesions showed All repairs were evaluated according to the ICRS complete tissue infill with good architectural protocol considering defect infill, repair tissue restoration and positive pericellular-specific pro- integration, and macroscopic appearance (Fig. 1A). teoglycan and type II collagen staining. Trochlea MACI-treated trochlea and condylar lesions were defect repair following untreated and bioscaffold significantly superior to controls at 10 weeks collagen-only intervention was characterized by (p < 0.05; Fig. 2A,B). Macroscopic analysis of con- fibrous tissue repair with partial infill, poor dylar lesion treatments demonstrated a high degree integration, poor architectural restoration, and of variability among treatment groups and time no proteoglycan or type II collagen production. points. Condylar defects treated with MACI and collagen- Laser Scanning Confocal Arthroscopy only bioscaffold showed initially poor regeneration that improved at 10 and 12 weeks but was not Intact hyaline/hyalinelike MACI-treated lesions significantly different to untreated controls were characterized by a high density of spatially (Fig. 2E). MACI-treated trochlear groups dis- uniform paired chondrocytes persisting in carti- played significant improvement in histological lage matrix (Fig. 1B). Some MACI and collagen- score in comparison to untreated controls at both only repairs showed a high density of non- 8 and 10 weeks (p < 0.05; Fig. 2F). uniformly distributed spindle-shaped cells within a collagenous fibril matrix. In defect-only controls, ulcerated tissue was either highly fibrous with a Magnetic Resonance Imaging low density of non-uniformly distributed cells or Improvement was seen in tissue infill and surface was largely acellular. Analysis of LSCA scores of continuity with MACI (Fig. 1D) compared to the trochlear cartilage repair tissue revealed a sig- collagen and untreated controls. MRI demon- nificant difference between the MACI and colla- strated superior results (p < 0.05) for MACI com- gen-only scaffold treatment groups at 10 weeks pared to both control and collagen groups at

DOI 10.1002/jor JOURNAL OF ORTHOPAEDIC RESEARCH 2007 Page 56

JONES ET AL.

Figure 1. (A) Macroscopic view of MACI-treated condyle. MACI graft was completely intact across entire defect with full-thickness infill achieved. Cartilage appeared pink in comparison to surrounding tissue. Area surrounding and immediately adjacent to MACI-treated region appears normal AC with no surface fibrillations or evidence of degenerative changes. (B) LSCA image of MACI repair tissue. Brightly stained round chondrocytes (C) persisted in high densities in the repair tissue. (C) Hematoxylin and eosin histology (original magnification, 100) of MACI-treated condyle with 100% tissue infill, excellent integration, and surface continuity. Note native AC (N) to repair tissue interface (MACI). Repair tissue was hyalinelike cartilage. (D) MRI image of MACI treatment (arrow indicates medial femoral condyle). Good integration of the graft was seen with smooth surface features, homogeneous consistency, and no edema. [Color figure can be viewed in the online issue, which is available at http://www.interscience.wiley.com.] all time points (Fig. 2G). No synovitis was recorded fold animals at 10 weeks). Following analysis of in any of the scanned animals. Due to limited initial MRI data, we concluded that assessment of availability of the MRI facility, two specimens trochlear defect sites was of limited value due were lost to follow-up (both collagen-only bioscaf- to the interference caused by a relatively high

JOURNAL OF ORTHOPAEDIC RESEARCH 2007 DOI 10.1002/jor Page 57

MACI IN SHEEP

Figure 2. Modality scores for each assessment modality at 8, 10, and 12 weeks after surgery. Data expressed as mean SD. Significant relationships (multiple pairwise Student t-test, p < 0.05) expressed for MACI versus other treatments. (A): Condyle ICRS modality scores showed a high degree of variability between treatment groups and time points. Condyle MACI repairs were significantly superior to controls at 10 weeks (p < 0.05). (B) Trochlea ICRS modality scores demonstrated the general superiority of MACI repairs, with significant superiority at 10 weeks (p < 0.05). (C) LSCA modality scores for condyle treatment groups showed neither MACI nor collagen-only scaffold groups demonstrated significant improvement in cartilage regeneration (p < 0.05). (D) LSCA modality scores for trochlea showed MACI was superior to the collagen-only scaffold treatment group at 10 weeks (p < 0.05). (E) Histological modality scores for condyle treatment groups showed neither MACI nor collagen-only scaffold treatments demonstrated significant improvement in cartilage regeneration (p < 0.05). (F) Histological modality scores for trochlea treatment groups showed MACI evidenced a significant improvement compared to untreated controls at both 8 and 10 weeks (p < 0.05). (G) MRI modality scores showed improvement for MACI condylar regeneration compared to collagen only and untreated controls at all three time points, with significance at 10 weeks (p < 0.05). signal-to-noise ratio and the difficulties of posi- between the MRI assessments and both the ICRS tioning the sample within the field. and LSCA modalities (Table 2). In assessment of the trochlea repair ORA grades, significant correlation was seen between the ICRS and the LSCA Overall Repair Assessment Correlation modalities (Table 2). No significant correlation was Pearson correlation analysis of condyle repair found between histology and other assessment ORA grades demonstrated significant correlations modalities.

DOI 10.1002/jor JOURNAL OF ORTHOPAEDIC RESEARCH 2007 Page 58

JONES ET AL.

Figure 3. Representative LSCA images with the corresponding modality score component breakdown provided in Supplementary Material (Table S1). (A) LSCA image of native articular cartilage demonstrating brightly stained classically paired chondrocytes (PC). LSCA score: 17. ORA grade I: normal. (B) LSCA image of MACI repair tissue from 10-week trochlear repair. High density round chondrocytes (C) persist in repair tissue. LSCA score: 15. ORA grade II: nearly normal. (C) LSCA image of degenerative tissue following creation of untreated control defect 10-week trochlear repair. A low density spindly shaped infiltrate (S) is evident deep in a necrotic lesion among whirls of connective tissue (W). LSCA score: 10. ORA grade III: abnormal. (D) LSCA image of fibrous tissue following spontaneous repair of untreated control defect 10-week trochlear repair. High densities of small spindle shaped fibroblasts (F) are evident. LSCA score: 7. ORA grade IV: severely abnormal.

JOURNAL OF ORTHOPAEDIC RESEARCH 2007 DOI 10.1002/jor Page 59

MACI IN SHEEP

Overall Treatment Effect Analysis Mechanical Testing Differences between mean modality scores for Data were obtained for 17 of 20 trochlear speci- treatment groups were assessed using the Krus- mens; the residual samples lacked the structural kal–Wallis test dependent and independent integrity required for testing. Samples were tested of time point. No signiﬁcant differences in overall independent of time point. All collagen-only bio- treatment effect were observed between time- scaffold treatment samples demonstrated load- point dependent treatment groups. Testing of bearing capacities below sensitivity of the load cell overall treatment effect independent of time point and were consequently removed from analysis. revealed signiﬁcant differences between MACI Data were obtained for 15 of 20 condylar speci- and control groups for all sites and assessment mens, however, once the collagen-only treatment modalities (Asymp Sig < 0.05) with the exception group was excluded, the low sample size and of the WAZ assessment of the condyle treatment highly variable nature of repair tissue resulted in groups (Asymp Sig ¼ 0.5; Fig. 4). data of limited quality. The stiffness relationship

Figure 4. Overall treatment effect independent of time point for (A) trochlea and (B) condyle treatment sites. A signiﬁcant improvement in mean modality score was seen across all assessment modalities for MACI treatment compared to collagen-only bioscaffold or control (except condyle histology). Kruskal–Wallis Asymp Sig values are presented. Data expressed as mean SD.

DOI 10.1002/jor JOURNAL OF ORTHOPAEDIC RESEARCH 2007 Page 60

JONES ET AL.

Figure 5. Lagrange stress versus relative vertical displacement. MACI represented by circles, control (fibrous) tissue by triangles, and native tissue by squares. [Color figure can be viewed in the online issue, which is available at http://www.interscience.wiley.com.] between the average vertical Lagrange stress and MACI repairs, compared to the moderate chon- the vertical displacement for the trochlear samples drocyte density evident in native tissue (Fig. 3). from the MACI and control treatment groups was Inconsistencies in repair quality between MACI- compared to the native tissue from corresponding treated trochlear and condylar lesions may have sites on the contralateral aspect of the trochlea been due to the influence of the loading environ- groove (Fig. 5). MACI samples showed a signifi- ment of the joint after surgery. Practical limita- cant degree of decreased stiffness in comparison to tions necessitated housing of animals in NHMRC- both the control and native tissue groups approved cages, thereby restricting postoperative (p < 0.05). ambulation, possibly accentuating static compression of the joint leading to a reduced matrix DISCUSSION production over time. Findings of our collagen-only bioscaffold group The regeneration of human articular cartilage are consistent with those of Frenkel and colleagues represents a major challenge.5 MACI is thought to and Dorotka and colleagues, who found that provide great promise for the treatment of complex unseeded collagen membrane implants are char- and multisite lesions and has enjoyed early clinical acterized by immature cartilage with poor archi- success.17,18 Large animal models of cartilage tectural restoration.21,23 Collagen-only treated repair have been conducted in dogs, pigs, goats, defects were inferior to MACI-treated defects, sheep, and horses.8,19–22 In a recent sheep study, a showing no significant improvement in overall multisite cartilage defect repair model demon- treatment effect compared to nontreated controls. strated that collagen matrix seeded with autolo- Interestingly, several collagen-only repairs dem- gous chondrocytes produced the best quantitative onstrated a low cellularity unreabsorbed mem- and qualitative results compared to microfrac- brane, the continued persistence of which remains ture.21 Similarly, in our study, analysis of the unclear. While perforation of the subchondral plate overall treatment effect independent of time point was avoided during implantation, many defects demonstrated a benefit for MACI in comparison to showed deterioration of the plate upon sacrifice. both collagen-only bioscaffold and controls (sig- Although collagen membrane has the ability to nificant across all assessment modalities except redifferentiate fibroblasticlike chondrocytes into condyle histology; Fig. 4) Cell density observations phenotypic chondrocytes, the cells seen in these from LSCA images revealed a vastly expanded defect repairs were of mixed morphology and chondrocyte population persisting in high quality displayed no significant matrix production upon

JOURNAL OF ORTHOPAEDIC RESEARCH 2007 DOI 10.1002/jor Page 61

MACI IN SHEEP immunohistochemistry.24 We did, however, observe confined configuration may be analyzed much the some chondrocytes within the residual fibrin seal- same as unconfined techniques. In contrast to the ant, a predictable response given previous results findings of Chu and colleagues,4 Briggs and suggesting that fibrin glue supports chondrocyte colleagues29 suggested that tissue-engineered car- growth.1,3 We believe the presence of cells in tilage may have inferior biomechanical properties membrane-only repair was a consequence of mar- compared to normal articular cartilage at early row deterioration through biomechanical factors stages of repair, due to a lack of complete architec- following defect creation with subsequent inflam- tural remodeling. Our results agree with these matory infiltration. Consistent with the results of findings, both mechanically and histologically. Dorotka and colleagues,21 we also observed the Testing of fibrous tissue retrieved from sponta- hallmarks of rare spontaneous repair of cartilage neously healed control defects closely resembled lesions within the defect-only control group. the native tissue in stiffness profile, and both were While a subjective clinical outcome is difficult to significantly superior to MACI tissue. Collagen- quantify in an animal model, our study compared only repairs were not of sufficient quality to tolerate nondestructive assessment techniques to conven- biomechanical testing. tional histology. The LSCA enabled high resolution While a beneficial treatment effect of MACI was microscopic imaging comparable to conventional demonstrated by ICRS, LSCA, and MRI assess- histology, but without causing tissue damage. ment, the results of histological analysis and Pearson correlation analysis of ORA grades dem- mechanical testing were less clear. Our results onstrated a significant degree of intermodality suggest that MACI is superior to controls for some agreement between LSCA, MRI, and ICRS overall measures but may be equal to or inferior to controls repair assessments. Consistent with Tins and for other measures. Long-term follow-up of the colleagues,9 no significant correlation was observed multisite MACI ovine model is therefore indicated. between MRI and histology (Table 2). Histological Interobserver analysis of randomized blinded scor- investigation also introduces trauma into a pre- ing of LSCA images has validated our scoring viously damaged joint and should ideally not be protocol with substantial to almost perfect agree- used for long-term human follow-up of cartilage ment. The development of nondestructive cartilage repair.5 Combined LSCA and MRI imaging may repair assessment using LSCA will facilitate the therefore provide a nondestructive method for long-term monitoring of focal cartilage defect assessing cartilage repair without the need for repair treatments. destructive biopsy. LSCA and MRI should be considered complementary imaging modalities, with LSCA providing cellular level resolution (axial ACKNOWLEDGMENTS 2 mm) and MRI providing a global image of the entire joint. Interobserver variability ICC analysis of LSCA Special thanks to Dr. Zeike Taylor for his biomechanical testing assistance. Thanks to the UWA large animal scores demonstrated substantial to almost perfect facility and to John Allen, Peter Delaney, and Wendy agreement between observers, thereby validating the McLaren of Optiscan. This work was supported by LSCA cartilage repair assessment scoring system grants from the Australian Research Council, SPRIT developed for this study. Prior to imaging in human Grant (Project ID: C00107367), and the National Health subjects, the toxicity of fluorophores should be more and Research Council Development Grant. The thoroughly explored; fluorescein is the only agent researchers are grateful for the equipment supplied by currently approved for direct systemic application Optiscan Pty Ltd. (Mt. Waverly, Victoria, Australia) and in human subjects, while acridine orange may Verigen Pty Ltd. (Leverkusen, Germany). inhibit mitosis and induce binucleation in chondrocytes. Regardless of histological quality, an important REFERENCES measure of cartilage repair lies in its biomechanical 5 1. Cherubino P, Grassi FA, Bulgheroni P, et al. 2003. properties. Previous cartilage compression tests Autologous chondrocyte implantation using a bilayer have utilized either confined or unconfined config- collagen membrane. J Orthop Surg 11. urations that may induce error into the modeling 2. Willers CR, Wood DJ, Zheng MH. 2003. A current review calculations.25–27 We employed a partially confined on the biology and treatment of articular cartilage defects configuration in which the cylindrical specimen (part I & part II). J Musculoskel Res 7:157–181. 3. Willers C, Chen J, Wood D, et al. 2005. Autologous sides were left unconstrained with the bounding 28 chondrocyte implantation with collagen bioscaffold for surfaces fixed to the load platens. Under the the treatment of osteochondral defects in rabbits. Tissue assumption of material anisotropy, the partially Eng 11:1065–1076.

DOI 10.1002/jor JOURNAL OF ORTHOPAEDIC RESEARCH 2007 Page 62

JONES ET AL.

4. Briggs TWR, Mahroof S, David LA, et al. 2003. Histological 17. Bartlett W, Skinner JA, Gooding CR, et al. 2005. evaluation of chondral defects after autologous chondro- Autologous chondrocyte implantation versus matrix- cyte implantation of the knee. J Bone Joint Surg [Br] 85: induced autologous chondrocyte implantation for osteo- 1077–1083. chondral defects of the knee: a prospective, randomised 5. Poole AR. 2003. What type of cartilage repair are we study. J Bone Joint Surg [Br] 87:640–645. attempting to attain? J Bone Joint Surg [Br] 85:40–44. 18. Gillogly SD. 2003. Treatment of large full-thickness 6. Brittberg M, Winalski CS. 2003. Evaluation of cartilage chondral defects of the knee with autologous chondrocyte injuries and repair. J Bone Joint Surg [Br] 85:58–68. implantation. Arthroscopy 19:147–153. 7. Richardson JB, Caterson B, Evans EH, et al. 1999. Repair 19. Chiang H, Kuo T, Tsai C, et al. 2005. Repair of porcine of human articular cartilage after implantation of auto- articular cartilage defect with autologous chondrocyte logous chondrocytes. J Bone Joint Surg [Br] 81:1064– transplantation. J Orthop Res 23:584–593. 1068. 20. Dell’Accio F, Vanlauwe J, Bellemans J, et al. 2003. 8. Breinan HA, Minas T, Hsu HP, et al. 2001. Autologous Expanded phenotypically stable chondrocytes persist in the chondrocyte implantation in a canine model: change in repair tissue and contribute to cartilage matrix formation composition of reparative tissue with time. J Orthop Res and structural integration in a goat model of autologous 19:482–492. chondrocyte implantation. J Orthop Res 21: 123–131. 9. Tins BJ, McCall IW, Takahashi T, et al. 2005. Autologous 21. Dorotka R, Windberger U, Macfelda K, et al. 2005. Repair chondrocyte implantation in knee joint: MR imaging and of articular cartilage defects treated by microfracture and a histologic features at 1-year follow-up. Radiology 234:501– three-dimensional collagen matrix. Biomaterials 26:3617– 508. 3629. 10. Jones CW, Keogh A, Smolinski D, et al. 2004. Histological 22. Nehrer S, Breinen HA, Ramappa A, et al. 1998. Chon- assessment of the chondral and connective tissues of drocyte-seeded collagen matrices implanted in a chondral the knee by confocal arthroscope. J Musculoskel Res 8: defect in a canine model. Biomaterials 19:2313–2328. 75–86. 23. Frenkel SR, Toolan B, Menche D, et al. 1997. Chondrocyte 11. Jones CW, Smolinski D, Wu JP, et al. 2004. Quantification transplantation using a collagen bilayer matrix for of chondrocyte morphology by confocal arthroscopy. J cartilage repair. J Bone Joint Surg [Br] 79:831–836. Musculoskel Res 8:145–155. 24. Fuss M, Ehlers EM, Russlies M, et al. 2000. Character- 12. Zheng MH, Willers CR, Kirilak L, et al. 2006. Matrix- istics of human chondrocytes, osteoblasts and fibroblasts induced autologous chondrocyte implantation (MACI1): seeded onto a type I/III collagen sponge under different biological and histological assessment. Tissue Eng 13:737– culture conditions. A light, scanning and transmission 746. microscopy study. Anat Anz 182:303–310. 13. Delaney PM, Harris MR, King RG. 1993. Novel microscopy 25. Armstrong CG, Lai WM, Mow VC. 1984. An analysis of the using fiber optic confocal imaging and its suitability for unconfined compression of articular cartilage. J Biomech subsurface blood vessel imaging in vivo. Clin Exp Phar- Eng 106:165–173. macol Physiol 20:197–198. 26. Buschmann MD, Soulhat J, Shirazi-Adl A, et al. 1998. 14. Recht M, Bobic V, Burstein D, et al. 2001. Magnetic reso- Confined compression of articular cartilage: linearity in nance imaging of articular cartilage. Clin Orthop S379– ramp and sinusoidal tests and the importance of inter- S396. digitation and incomplete confinement. J Biomech 31:171– 15. Marlovits S, Singer P, Zeller P, et al. 2006. Magnetic 178. resonance observation of cartilage repair tissue (MOCART) 27. Jurvelin J, Buschmann MD, Hunziker EB. 2003. Mechan- for the evaluation of autologous chondrocyte transplantation: ical anisotropy of the human knee articular cartilage in determination of interobserver variability and correlation to compression. J Eng Med 217:215–219. clinical outcome after 2 years. Eur J Radiol 57:16–23. 28. Miller K. 2005. Method of testing very soft biological 16. McGraw KO, Wong SP. 1996. Forming inferences tissues in compression. J Biomech 38:153–158. about some intraclass correlation coefficients [erratum 29. Chu CR, Dounchis JS, Yoshioka M, et al. 1997. Osteochon- published in Psychol Method 1:390]. Psychol Method 1: dral repair using perichondrial cells.A 1-year study in 30–46. rabbits. Clin Ortho p 220–229.

JOURNAL OF ORTHOPAEDIC RESEARCH 2007 DOI 10.1002/jor

Chapter 5

Clinical MACI: Biology and Histology

Thesis publication #4: Zheng MH, Willers C, Kirilak L, Yates P, Xu J, Wood D, Shimmin A. Matrix-induced autologous chondrocyte implantation (MACI): biological and histological assessment. Tissue Eng. 2007; 13(4): 737-46.

Page 63

STATEMENTS OF AUTHORSHIP CONTRIBUTION FROM CO-AUTHORS

Matrix-induced autologous chondrocyte implantation (MACI): biological and histological assessment.

Zheng MH, Willers C, Kirilak L, Yates P, Xu J, Wood D, Shimmin A.

Published in the Tissue Engineering 2007, volume 13(4), pages 737-746.

Zheng MH (Supervisor) Major contribution to planning of research. Moderate contribution to execution of research (collected all biopsies). Major contribution to analysis and interpretation of research data. Moderate contribution to writing of manuscript.

Signature of Principal Author: ……………………………… Date: .………………

Willers C (PhD Candidate) Major contribution to execution of research (histology and SEM). Major contribution to analysis and interpretation of research data. Major contribution to writing of manuscript.

Signature of Co-Author: ……………………..……………… Date: …………………

Kirilak Y (Research Collaborator) Major contribution to execution of research (histology and immunohistochemistry). Moderate contribution to analysis of research data.

Signature of Co-Author: ……………………..……………… Date: …………………

Page 64

Yates P (Research Collaborator) Minor contribution to research planning. Minor contribution to execution of research.

Signature of Co-Author:……………………..………………Date:…………………

Xu J (Research Collaborator) Minor contribution to execution of research. Reviewed manuscript prior to submission.

Signature of Co-Author:……………………..………………Date:…………………

Wood D (Co-Supervisor) Minor contribution to execution of research. Reviewed manuscript prior to submission.

Signature of Co-Author:……………………..………………Date:…………………

Shimmin A (Research Collaborator) Minor contribution to execution of research. Reviewed manuscript prior to submission.

Signature of Co-Author:……………………..………………Date:…………………

Page 65

STATEMENTS OF PERMISSION FROM CO-AUTHORS

I, Ming-Hao Zheng give permission to Craig Willers, second author of the paper: Matrix-induced autologous chondrocyte implantation (MACI): biological and histological assessment., published in Tissue Engineering (2007), to include this paper as Chapter 5 of his PhD thesis entitled Matrix-induced Autologous Chondrocyte Implantation for Articular Cartilage Injury: Biology, Histology, and Clinical Outcomes.

Signature: …………………………………. Date: ………………………………….

I, Yaowanuj Kirilak give permission to Craig Willers, second author of the paper: Matrix-induced autologous chondrocyte implantation (MACI): biological and histological assessment., published in Tissue Engineering (2007), to include this paper as Chapter 5 of his PhD thesis entitled Matrix-induced Autologous Chondrocyte Implantation for Articular Cartilage Injury: Biology, Histology, and Clinical Outcomes.

Signature: …………………………………. Date: ………………………………….

I, Piers Yates give permission to Craig Willers, second author of the paper: Matrix- induced autologous chondrocyte implantation (MACI): biological and histological assessment., published in Tissue Engineering (2007), to include this paper as Chapter 5 of his PhD thesis entitled Matrix-induced Autologous Chondrocyte Implantation for Articular Cartilage Injury: Biology, Histology, and Clinical Outcomes.

Signature: …………………………………. Date: ………………………………….

Page 66

I, Jiake Xu give permission to Craig Willers, second author of the paper: Matrix- induced autologous chondrocyte implantation (MACI): biological and histological assessment., published in Tissue Engineering (2007), to include this paper as Chapter 5 of his PhD thesis entitled Matrix-induced Autologous Chondrocyte Implantation for Articular Cartilage Injury: Biology, Histology, and Clinical Outcomes.

Signature: …………………………………. Date: ………………………………….

I, David Wood give permission to Craig Willers, second author of the paper: Matrix- induced autologous chondrocyte implantation (MACI): biological and histological assessment., published in Tissue Engineering (2007), to include this paper as Chapter 5 of his PhD thesis entitled Matrix-induced Autologous Chondrocyte Implantation for Articular Cartilage Injury: Biology, Histology, and Clinical Outcomes.

Signature: …………………………………. Date: ………………………………….

I, Andrew Shimmin give permission to Craig Willers, second author of the paper: Matrix-induced autologous chondrocyte implantation (MACI): biological and histological assessment., published in Tissue Engineering (2007), to include this paper as Chapter 5 of his PhD thesis entitled Matrix-induced Autologous Chondrocyte Implantation for Articular Cartilage Injury: Biology, Histology, and Clinical Outcomes.

Signature: …………………………………. Date: ………………………………….

Page 67

TISSUE ENGINEERING Volume 13, Number 4, 2007 # Mary Ann Liebert, Inc. DOI: 10.1089/ten.2006.0246

Matrix-Induced Autologous Chondrocyte Implantation (MACIÒ): Biological and Histological Assessment

MING-HAO ZHENG, Ph.D., D.M., FRCPath,1 CRAIG WILLERS, B.Sc.(H1), M.Med.Sc.,1 LYN KIRILAK, M.Sc.,1 PIERS YATES, MBBS (Hons), B.Sc. (Hons), MRCS, FRCS,1 JIAKE XU, MBBS, Ph.D.,1 DAVID WOOD, B.Sc., MBBS, M.S., FRCS, FRACS,1 and ANDREW SHIMMIN, MBBS, FAOrthA, Dip. Anat.2

ABSTRACT

Matrix-induced autologous chondrocyte implantation (MACIÒ) has been a treatment of cartilage injury since 2000, but little is known of the histological paradigm of tissue regeneration after implantation. MACI is a stable cell-based delivery system that enables the regeneration of hyaline-like cartilage. From a cohort of 56 MACI patients, we examined the phenotype of chondrocytes seeded on type I/III collagen scaffold, and conducted progressive histologic assessment over a period of 6 months. Chondrocyte-seeded collagen scaffolds from patient implants were analyzed by electron microscopy, immunohistochemistry (type II collagen and S-100), and reverse transcription polymerase chain reaction (RT-PCR) (aggrecan and type II collagen). Coincidental cartilage biopsies were obtained at 48 hours, 21 days, 6 months, 8 months, 12 months, 18 months, and 24 months. Our data showed that chondrocytes on the collagen scaffold appeared spherical, well integrated into the matrix, and maintained the chondrocyte phenotype as evidenced by aggrecan, type II collagen, and S-100 expression. Progressive histologic evaluation of the biopsies showed the formation of cartilage-like tissue as early as 21 days, and 75% hyaline-like cartilage regeneration after 6 months. This preliminary study has suggested that MACI may offer an improved alternative to traditional treatments for cartilage injury by regenerating hyaline-like cartilage as early as 6 months after surgery.

INTRODUCTION gous chondrocytes under a periosteal flap placedover chondral defects is capable of regenerating articular cartilage of rabbit 13 ARTILAGE INJURY IS irreversible within current treatment patella. Implantation of human chondrocytes achieved 87% Cparameters. Arthroscopic lavage, abrasion arthropla- good and excellent clinical results in 24 months, and demon- sty, subchondral drilling, microfracture, and osteochondral strated 73% hyaline-like cartilage regeneration in femoral autografting/allografting are used to treat articular cartilage condylar injury.12 However, the large surgical incision, peri- injury, but result in predominantly fibrocartilage repair ra- pheral graft hypertrophy, graft delamination, and possible ther than hyaline cartilage regeneration, thus causing variable calcification associated with the use of periosteal flap have clinical outcomes.1–8 The shortcomings of these approaches hindered the efficacy of ACI.17,18,22–26 Postoperative experi- have stimulated the development of alternative strategies ence suggests that a significant percentage (20%) of patients such as autologous chondrocyte implantation (ACI).9–21 present with symptomatic catching of the joint because of Peterson et al. showed the first crucial evidence that ACI hypertrophic periosteal graft edges which lead to the need for was a therapeutically credible treatment for chondral in- revision arthroscopy.18 Poor integration of the periosteal flap jury.12,13,20 It has been shown that transplantation of autolo- and periosteal delamination has been reported, and it has been

1Department of Orthopaedics, School of Pathology and Surgery, University of Western Australia, Nedlands, Perth, Australia. 2Melbourne Orthopaedic Group, The Avenue Hospital, Windsor, Melbourne, Australia. Page 68

ZHENG ET AL. confirmed that periosteum can form ectopic bone.22,23,27 In collagen scaffolds. Hematoxylin and eosin histology, immu- general, periosteal complications after ACI could occur in at nohistochemistry of type II collagen, and biomechanical stiff- least 25% of patients.18 ness measurement were used to examine the progression of Complications arising from the use of periosteal flap have regenerative cartilage tissue. stimulated the development of alternate scaffolds for ACI, Type I/III collagen scaffolds from ACI-MaixÔ (Matricel including Hyalograft and type I/III collagen membrane.27,28 GmbH, Herzogenrath, Germany) were used for MACI chon- The use of alternative scaffolds instead of periosteum avoids drocyte implantation. The procedures were conducted be- the need to microsuture the defect and the unnecessary tween four Australian private hospitals by four orthopedic disruption to the adjacent cartilage. Matrix-induced autolo- consultants (David Wood, Andrew Shimmin, Keith Holt, gous chondrocyte implantation (MACIÒ) uses type I/III Tony Ganko). All patients had previously undergone three or collagen membrane and TisseelÒ fibrin sealant to attach the more unsuccessful knee operations. Inclusion was based on chondrocyte-seeded scaffold and fill the cartilaginous de- the following: patient age from 15 to 55 years, signed con- fect void.6,14,29 Six-month postoperative results suggested sent, patient understood and followed rehabilitation protocol, improved clinical and functional outcomes, MRI-visualized a focal unipolar defect classified as Outerbridge Grade IV33 hyaline-like cartilage, and no complications.14 Moreover, (a defect no larger than 15 cm2 after debridement), no history hyaline-like repair, improved clinical outcome, and reduced of gentamicin sensitivity, and a biomechanically stable knee graft hypertrophy were seen with the fibrin sealant method (negligible joint laxity detected). Exclusion criteria were in comparison to collagen-covered ACI.30 However, whilst based on diagnosis of osteoarthritis, rheumatoid arthritis, the observed improvements in clinical outcomes are prom- chondrocalcinosis, malalignment > 58 varus or valgus, patel- ising, the therapeutic efficacy of this technology needs to be lofemoral instability, and obese patients (body mass index further elucidated. In particular, whilst it has previously [BMI] > 30). Age of patients who had a biopsy ranged from been demonstrated that cultivated chondrocytes have a low 20 to 51 years (mean: 39.9 11.73 years) at the time of sur- apoptotic index prior to and upon membrane seeding, main- gery. There was a gender distribution of six women to four tenance of the seeded chondrocyte phenotype and the bio- men (Table 1). Defects ranged from 1 to 14 cm2 (mean: logical paradigm of cartilage repair immediately following 5.32 cm2 4.30 cm2) in area, were all of full thickness (Grade MACI are not documented.31,32 Although it is certainly pos- IV), and all except one (patella) were located on the medial sible that clinical results could be independent of objective femoral condyle of the knee. parameters such as preimplantation cell phenotype and The surgical procedure of MACI began with the arthro- histological repair progression, analysis of these variables scopic harvesting of cartilage tissue for chondrocyte culti- may help determine and optimize the nature of regeneration vation. Cartilage tissue (50–150 mg) was obtained from the following surgery. nonweightbearing supracondylar region about the femoral In this preliminary study, we have examined the char- condyles and placed into serum-free nutrient media.31 The acteristic features of type I/III collagen membrane before isolation and cultivation of autologous chondrocytes was and after autologous chondrocyte seeding, and the timeline conducted in Australian Therapeutic Goods Administration of cartilage regeneration following MACI intervention. (TGA), licensed facilities at Verigen Australia, as according to the method described by Brittberg et al.12 After acceptable cell density was ascertained in vitro, cells were MATERIALS AND METHODS seeded onto the collagen membrane and transported to the- ater. The surgical technique used was similar to that de- We conducted a retrospective study of 56 patients treated scribed by Cherubino et al.14 The defect site was accessed by MACI. Within the cohort, only 11 patients consented to via a parapatellar incision and arthrotomy in a tourniquet biopsy of their repair cartilage for histological assessment. controlled field. During implantation, defects were thor- The biopsy numbers were limited because of patients’ re- oughly curetted to remove reactive fibrous tissue build-up luctance to undergo further intervention not related to their and define defect borders. The chondrocyte-seeded colla- treatment. All patients consented to having MACI, and eth- gen scaffold was then shaped to match defect geometry. ical approval was given by the Human Research Ethics Com- Bleeding was controlled by adrenalin swab. Once the scaf- mittee of the University of Western Australia. The experiment folds were correctly shaped, the defect was filled to the was conducted in two parts. Part 1 was designed to evaluate surface by the injection of Tisseel fibrin sealant (Baxter the biological features of implanted chondrocytes and porcine- AG, Vienna, Austria) and the shaped scaffold was press-fit derived type I/III collagen scaffold; part 2 was aimed at the into the defect. Full-range movement of the joint was made progressive analysis of regenerative tissue after MACI. Elec- (three to four times) before closure to assure implant tron microscopy, immunohistochemistry of S-100 and type II stability. Continuous passive motion was commenced 1 day collagen, and reverse transcription polymerase chain reaction postoperatively, and the patients were gradually returned to (RT-PCR) analysis of aggrecan and type II collagen were weightbearing activity over the ensuing months by partic- used to analyze the microstructure and cellular characteristics ipation in a graduated rehabilitation program designed for of autologous chondrocytes from patients and the implanted ACI recovery.34 Page 69

MATRIX-INDUCED AUTOLOGOUS CHONDROCYTE IMPLANTATION

Table 1. BIOPSIED CASE INFORMATION

Patient Age (years) Gender Defect size (cm2) Defect site Cartilage (mg) Viability (%) Seeding density

1 48 Female 2 MFC 60 99 2.0107 2 47 Female 14 MFC 82 97 9.9106 3 20 Male 2.25 MFC 1000 97 1.6107 4 36 Male 4 MFC 190 98 9.0106 5 20 Male 6 MFC 185 83.5 2.36107 6 51 Female 1 MFC 60 99 1.06107 7 36 Female 4 MFC 154 98 1.03107 8 48 Female 12 Patella 260 98 2.06107 9 43 Female 4 MFC 43 99 1.51107 10 60 Male 8 MFC 66 92.3 1.95107 11 50 Male 4 MFC 135 98 1.29107

MFC ¼ medial femoral condyle

We conducted coincidental biopsies at follow-up arthros- osmium tetroxide in cacodylate buffer for 60 minutes at copy or open surgery (because of concomitant surgical pro- room temperature; washed in three changes of cacodylate cedures) to assess the histological features of cartilage tissue buffer; placed in 1% tannic acid in 0.05 M cacodylate buffer after implantation. At arthroscopy, 2 mm core biopsies were for 60 minutes at room temperature; washed in saline solu- taken from a representative area in the center and/or peri- tion; stained for 60 minutes in 0.5% uranyl acetate in double- phery of the implanted area. Implanted areas have an obvious distilled water (DDW); rinsed well in saline solution; placed border in arthroscopic visualization, making biopsy location in 25%, 50%, 70%, 95%, and absolute ethanol sequentially accurate. Biopsy conducted at 48 hours was because of the for 30 minutes each at room temperature; then washed twice recall of the procedure (n ¼ 1). Biopsy conducted at 21 days in super-dry ethanol for 30 minutes at room temperature. was because of postoperative infection (the implant was not For TEM, samples were embedded in araldite and cut into affected; n ¼ 1). Biopsy at 6 months was performed in a ultrathin (50 nm) sections following dehydration. After crit- workers’ compensation patient who complained of no im- ical point drying, samples were mounted and viewed using provement (n ¼ 1). One biopsy at 8 months and two biopsies a Phillips XL30 (Phillips, Eindhoven, The Netherlands) scan- at 12 months were performed in a patient to acquire justifi- ning electron microscope or Phillips 600 transmission elec- cation on the performance of maximum physical activity. tron microscope. Two 12-month and one 18-month biopsies were performed To test the maintenance of the chondrocyte phenotype because other surgical procedures to the joint were required. within the scaffold, chondrocytes derived from in vitro One complete MACI graft was harvested from a patient who monolayer culture and chondrocyte-seeded scaffold from the died unexpectedly by unrelated means at 18 months after same patients were subjected to RT-PCR. After an equal MACI. The 24-month biopsy was performed because a sec- growth period (4 days), cells from the collagen scaffold and ond MACI procedure was introduced for other defects. The monolayer cultures were lysed and total RNA was extracted biopsies were placed into 4% paraformaldehyde immediately using RNA-Bee (Tel-Test, Friendswood, TX) according to after surgery. The tissues were decalcified with 10% formic the manufacturer’s instructions. Total RNA was then sub- acid, dehydrated by a graded series of alcohol and xylene jected to RT-PCR as previously described.35 Polymerase washes, and paraffin embedded. Histological outcomes were chain reaction (PCR) was achieved using 1.0 U of Taq poly- characterized by a qualified pathologist (M.H. Zheng) after merase (Boehringer Mannheim, Mannheim, Germany); 2 mL histological and immunohistochemical processing of the bi- of primers specific to human aggrecan, type II collagen, and opsies. Because of the lack of concurrent control biopsies, glyceraldehyde-3-phosphatedehydrogenase(GAPDH;20 mM); historical patient controls from the literature were utilized for 125 mMofdNTPin1PCR buffer (Boehringer Mannheim); comparison. and DDW in a total volume of 25 mL. Primers were used based To evaluate the morphological features of chondrocyte- on their published GenBank sequence31 listed as aggrecan: seeded and -unseeded collagen scaffold, scanning electron sense 50-GCATTCTGGATTTCTGGACC-30, antisense 50-AG microscopy (SEM) and transmission electron microscopy GTTAGCTTCGTGGAATGC-30; type II collagen: sense 50- (TEM) were used to assess the ultrastructure of the scaffold GTCATTTCCTTGTGCTCTCC-30, antisense 50-ATGGGCA and chondrocyte attachment in vitro. Surplus chondrocyte- GCAGTGTTTCTCC-30; GAPDH: sense 50-GGAGTCAACG seeded scaffold from leftover implantation was used for the GATTTGGT-30, antisense 50-GTGATGGGATTTCCATTGA evaluation. The seeded (4-day culture) and unseeded scaf- T-30. The amplification was performed in a DNA thermal folds were fixed in 2.5% glutaraldehyde for 7 days at room cycler (model 2400; PerkinElmer, Boston, MA). Polymerase temperature and treated with tannic acid. Samples were chain reaction conditions consisted of 35 cycles at 948Cfor rinsed well in 0.2 M cacodylate buffer; postfixed in 1% 5 minutes, with annealing conducted at 588C, 588C, and 558C Page 70

ZHENG ET AL. for 1 minute for aggrecan, collagen II, and GAPDH, respec- on the smooth surface are seen aggregating together to tively, and extension at 728C for 1 minute. The predicted PCR provide a slick surface reminiscent of the articular cartilage product sizes for aggrecan, type II collagen, and GAPDH were surface (Fig. 1B). In contrast, the rough surface is charac- 492, 384, and 206 bp, respectively. Because of the priority terized by its loose, porous collagen fiber arrangement ca- given to chondrocyte implantation and subsequent limitations pable of chondrocyte inoculation and integration (Fig. 1C). on cell number, only selected patients with surplus chondro- Scanning electron microscopy of the chondrocyte-seeded cyte growth were subjected to RT-PCR. scaffold showed the integration and attachment of chondro- For immunohistochemistry of S-100 and type II collagen cytes within the collagen matrix of the scaffold (Fig. 1D, E), of biopsied tissue and chondrocyte-seeded collagen scaffold, as well as their differentiated globular morphology (Fig. 1E). sections were deparaffinated with xylene, rehydrated with de- Additionally, TEM further demonstrated the anchoring of creasing ethanol solutions, and rinsed for 10 minutes in DDW. chondrocytes to the scaffold collagen fibers via cytoplasmic Sections were microwave heated in pH 6 citrate buffer, five projections (Fig. 1F). times for 2 minutes at low power for antigen retrieval. Sec- The preservation of chondrocytic phenotype by the type tionswerethenrinsedin3%hydrogenperoxideinmethanol I/III collagen scaffold was evidenced by the presence of for 15 minutes to block endogenous peroxidase activity. Non- S-100 and type II collagen positive cells within the scaffold, specific staining was blocked by rinsing sections in 10% fetal and the expression of type II collagen and aggrecan mRNA bovine serum (FBS) in 0.1% triton X-100 in tris buffered by the cells after membrane seeding (Fig. 2A–D). Under low saline (TBS) for 30 minutes before incubation for 2 hours magnification, S-100 positive cells were well integrated with primary monoclonal antihuman (1:100) type II collagen within the matrix of the collagen scaffold (Fig. 2A). These mouse IgG (ICN Biomedicals, Chemical Credential, Irvine, cells were also positive for type II collagen staining before CA) or S-100 protein (Zymed, South San Francisco, CA) implantation (Fig. 2B). The typical cell density of seeding diluted in 0.1% triton X-100/1% bovine serum albumin (BSA) was more than 5106/cm2. Similar immunohistochemical in TBS. After washing three times for 5 minutes in TBS, results were obtained from all patients tested. Analysis of 30 sections were incubated with EnVision þ goat anti-mouse implanted membranes showed that 80% of patients had 80% IgG peroxidase system at room temperature for 30 minutes. or greater cells positive for S-100 protein, whilst 50% of Sections were washed three times for 5 minutes in TBS, and patients had 80% or greater cells positive for type II collagen then incubated in 2% liquid diaminobenzidine (DAB) for (Fig. 2C). RT-PCR further confirmed the suitability of the 7 minutes. Sections were washed in distilled water, counter- scaffold for MACI through the phenotypic expression of stained shortly with hematoxylin, washed, stained shortly aggrecan and type II collagen in the autologous chondro- with Scott’s tap water, washed in distilled water, then de- cytes after seeding onto the type I/III collagen scaffold (Fig. hydrated, cleared, and mounted. Sections were viewed under 2D). While control GAPDH showed similar band intensity light and polarized microscope. All histological assess- in both samples, aggrecan and type II collagen intensities for ments were made by a consultant pathologist (M.H. Zheng) the seeded chondrocytes were higher than monolayer cul- blinded to the results. Thirty patients from the cohort had ture. leftover membrane from their implantation assessed for The sequential biopsy of cartilage tissue from 48 hours percentage positivity of type II collagen and S-100 protein. to 24 months after MACI treatment showed a steady pro- Positivity was assessed by counting the percentage of posi- gression of cartilage-like tissue regeneration at 21 days, to tive-staining cells in the total cell population across five fields hyaline-like cartilage formation as early as 6 months postop- (200magnification) toward the center of the membrane. eratively (Table 2; Figs. 3–6). At 48 hours, the entire repair Mean percentage was calculated per patient and classified tissue was observed to be a mix of spindle-shaped to round as > 80%, 50–80%, < 50%, or 0% positivity. cells scattered amongst their fibrin glue housing, with no obvious evidence of cartilage matrix formation (Fig. 3A). Most cells migrated out of the collagen membrane and scat- RESULTS tered into the fibrin glue matrix. After 21 days, there was a heterogeneous mix of cartilage-like matrix and mesenchymal Scanning electron microscopy assessment of the type tissue throughout the defect (Fig. 3B). Cells located within I/III collagen scaffold showed a bilayer collagen fiber orga- foci of cartilage-like matrix appeared round and resembled nization that maintained spherical chondrocyte morphology, chondrocytes (Fig. 3B), whereas cells within the mesen- and facilitated cell integration and matrix synthesis upon chymal tissue matrix appeared more elongated and con- chondrocyte seeding (Fig. 1A–F). The approximate dry tained large nuclei characteristic of mesenchymal cells. At 6 thickness of the scaffold was 400 mm. Cross-sectional im- months, a biopsy showed hyaline-like cartilage morphologic aging of the collagen scaffold showed the compact collagen features, with chondrocytes within matured lacunae and zonal fiber arrangement on the smooth surface of the membrane, cellular organization across the tissue (Fig. 3C, D). Cellular compared to the loose arrangement of fibers on the rough morphology of chondrocytes within the matrix ranged from surface (Fig. 1A). The smooth scaffold surface shows a cell- spherical to spindle shaped (Fig. 3C). Also, chondrocytes were occlusive compact arrangement of collagen fibers. Fibers noted to be at high density within the type II collagen positive Page 71

MATRIX-INDUCED AUTOLOGOUS CHONDROCYTE IMPLANTATION

100 90 80 70 60 50 40 % Patients 30 20 10 0 S-100 Type II Collagen

GADPH Collagen II Aggrecan

500 bp

250 bp

12345 67

FIG. 2. Phenotypic analysis of chondrocytes following seeding onto ACI-Maix bilayer collagen scaffold. (A) S-100 positive FIG. 1. Morphological analysis of chondrocyte seeding onto staining shows that the integrated cells within the collagen matrix ACI-Maix bilayer collagen scaffold by electron microscopy. (A) are chondrolineage, and shows their distribution throughout the Cross-sectional SEM imaging of the scaffold shows the differential rough surface of the scaffold (original magnification, 20). (B) organization of the collagen matrix. (B) SEM imaging of the cell- Staining of the chondrocyte-seeded scaffold for type II collagen occlusive compact arrangement of collagen in the smooth surface. was positive, which suggests chondrolineage cell presence and (C) The rough surface of the collagen scaffold showing its loose indicates that the synthesis of collagenous matrix in the seeded collagen matrix within the scaffold under SEM. (D) SEM imaging scaffold is active prior to implantation (original magnification, of the chondrocyte-seeded scaffold shows the differentiated glob- 20). (C) Analysis of 30 patients for percentage S-100 and col- ular chondrocyte appearance and attachment of chondrocytes to the lagen II positivity evidenced that 80% of patients had 80% or collagen fibers of the scaffold. (E) High magnification SEM of an greater cells positive for S-100 protein, whilst 50% of patients had individual chondrocyte seeded onto the collagen scaffold shows 80% or greater cells positive for type II collagen. Interestingly, 13% cell attachment via cytoplasmic (philopodia) projections (arrow) of assessed patients showed no evidence of type II collagen upon after seeding. (F) TEM imaging of chondrocyte attachment with staining. (D) RT-PCR comparison shows that human chondrocyte the scaffold showing the presence of cytoplasmic projections an- aggrecan and type II collagen gene expression of the same patient choring cells to the collagen fibers of the scaffold (magnification, is maintained from monolayer culture to cell inoculation onto the 15,000 ). collagen scaffold. Aggrecan and type II collagen expression seemed slightly higher in the seeded chondrocytes compared to monolayer culture. Lane 1 represents molecular weight standard bands; Lanes 2, matrix (Fig. 3D). Hyaline to hyaline-like cartilage tissue was 4, and 6 represent chondrocytes cultured in monolayer conditions; observed at 8 months (Fig. 4A–C). Chondrocytes tended to and Lanes 3, 5, and 7 represent chondrocytes seeded onto ACI-Maix distribute as columnar arrangements or clusters within the scaffold. matrix, but arranged mainly as spindle-shaped cells in the superficial zone and round chondrocytes within their lacunae in the deeper zones (Fig. 4B). Collagen fibrils, possibly de- two patients showed uniformly characteristic hyaline-like rived from the scaffold, were still obvious in the intermediate cartilage regeneration similar to the 8-month biopsy histology. zone, but the matrix produced by chondrocytes in the deep The superficial zone of the regenerated cartilage in both pa- zone was mainly well-integrated hyaline-like cartilage. It was tients contained elongated chondrocytes and displayed a very noteworthy that a small amount of fibrin glue remained within smooth surface similar to 8-month biopsy histology (Fig. 5A). the matrix of the transitional cartilage zone in this one patient Distribution of chondrocytes in both patients was a mix of (Fig. 4C). At 12 months, one of four biopsies showed fibro- columnar organization and clusters. Type II collagen matrix cartilage tissue. Of the remaining three, one biopsy showed production at 12 months (Fig. 5B) was also similar to healthy hyaline-like cartilage with foci of fibrocartilage (Fig. 5A), and hyaline articular cartilage. In one of three patients, collagen Page 72

ZHENG ET AL.

FIG. 4. Histology micrographs from 8 months postoperatively show the predominance of hyaline-like regeneration. (A) Low FIG. 3. Histology of MACI-induced cartilage repair from magnification of the biopsy tissue shows the hyaline-like appear- 48 hours to 6 months postoperatively. (A) Histology micrograph of ance of the regenerative cartilage, with spindle-shaped cells in the the transitional zone of the regenerated tissue at 48 hours shows superficial zone and rounded cells in the transitional and radial the spherical morphology of implanted autologous chondrocytes zones (Stain, hematoxylin and eosin; original magnification, 20). within the fibrin glue (Stain, hematoxylin and eosin; original mag- (B) Globular chondrocytes within their matrix lacunae are seen in nification, 100). (B) The 21-day repair photomicrograph shows a the transitional zone of the regenerated cartilage, illustrating the heterogeneous mix of spherical chondrocytes and cartilage-like maturation of tissue architecture. Collagen fibrils (arrowhead) matrix within mesenchymal tissue. Fibrin glue is shown within the possibly derived from the scaffold. (Stain, hematoxylin and eosin; transitional zone of the regenerated tissue at 21 days (Stain, hema- original magnification, 200). (C) Some residual fibrin sealant toxylin and eosin; original magnification, 100). (C, D) Histology was observed in the transitional zone of the regenerative carti- of hyaline-like regeneration is shown at 6 months postoperatively. lage (Stain, hematoxylin and eosin; original magnification, 200). (C) Hyaline-like cartilage was seen in the superficial zone of the regenerated cartilage with spindle-shaped chondrocytes, mature chondrocytes within their hyaline-like matrix lacunae, and zonal organization similar to a healthy state (Stain, hematoxylin and of proteoglycan staining (not shown), and surface fibrillation eosin; original magnification, 200). (D) Type II collagen immu- (Fig. 6B). The MACI graft was characterized by a cartilagi- nohistochemistry 6 months after collagen scaffold ACI demonstrated strong positive staining in the radial zone reminiscent of nous tissue (Fig. 6C, D) containing abundant proteoglycan native tissue (original magnification, 200). (Fig. 6E) and type II collagen (not shown), round-shaped chondrocytes located within lacunae, and chondrocyte density that was higher than in the native cartilage. The histologi- fibrils and elastin fibers from the type I/III collagen scaffold cal characteristics of this MACI graft are representative of were evidenced in the cartilage matrix, but no inflammatory hyaline-like cartilage.11 However, eburnation was noted at the response or lymphocytic infiltration was observed. Eighteen- inferior extremities of the grafted site (Fig. 6A). month histology and type II collagen staining were identical to 12-month histology, but no remaining scaffold fibrils were observed. At 24 months, this biopsy showed excellent DISCUSSION hyaline-like cartilage regeneration (Fig. 5C). Although there was good integration with the surrounding host cartilage and Autologous chondrocyte implantation using a periosteal higher cell density, there was minor surface separation be- flap was first to highlight the therapeutic potential of au- tween the host cartilage and the regenerative cartilage tologous cell techniques, but various complications have (Fig. 5D). been reported.12,17,22,23,25,30,36 Hence, the development of Interestingly, histological analysis of the MACI graft re- collagen scaffold instead of periosteum to seal in chondro- trieved at 18 months illustrated hyaline-like cartilage regen- cyte injections has been welcomed for its ability to reduce eration across the whole graft (Fig. 6). The adjacent native tissue hypertrophy, decrease surgical invasiveness, reduce articular cartilage appeared to be yellowish as compared to operating time, minimize donor site morbidity and post- the MACI graft (Fig. 6A) and showed the histological fea- operative pain, and avoid complications.11,14,30 MACI uses tures of arthritic degeneration, including cell clustering, loss type I/III collagen scaffold in conjunction with fibrin sealant Page 73

MATRIX-INDUCED AUTOLOGOUS CHONDROCYTE IMPLANTATION

FIG. 5. Histology of MACI-induced cartilage repair from 12 to 24 months postoperatively. (A) Biopsy histology micrograph from 12 months postoperatively shows the persistence of hyaline-like cartilage in the regenerated cartilage. Higher magnification of the superficial zone exhibits the smooth surface and superficial spindle cellular morphology of the regenerated hyaline-like cartilage (Stain, hematoxylin and eosin; original magnification, 100). (B) Type II collagen immunohistochemistry is shown 12 months after collagen scaffold ACI. Note the strong positive staining in the radial zone reminiscent of native tissue (original magnification, 100). (C) A histology micrograph from 24 months postoperatively FIG. 6. Histology and appearance of a retrieved MACI graft shows the persistence of hyaline-like cartilage in the regenerated after 18 months repair. (A) The MACI-grafted defect appeared cartilage. Hyaline-like morphology of the regenerated cartilage is white and smooth in texture macroscopically, as seen with healthy seen in the superficial and transitional zones, with moderate cell cartilage, whereas the surrounding knee surface was yellow and density, spherical chondrocytes within lacunae, and no obvious cel- appeared fibrillated as with osteoarthritic cartilage (photo taken lular architecture (Stain, hematoxylin and eosin; original magnifi- after slicing for sectioning). (B) Osteoarthritis of the cartilage cation, 100). (D) Histology of the healthy regenerative tissue immediately adjacent to the MACI-grafted defect (imaged from interface 24 months after MACI. Histology between the healthy (H) center of the defect—dashed line in panel A), with characteristic and regenerative (R) cartilage at 24 months evidenced a cleft at the fibrillation of the articular cartilage surface, reduced cell density defect interface. The regenerative cartilage showed higher cell den- and chondrocyte clustering (Stain, hematoxylin and eosin; original sity than the abutting host cartilage (Stain, hematoxylin and eosin; magnification, 40). (C) Low magnification of the MACI regen- original magnification, 40). erate (dashed line in panel A) shows the homogeneous matrix appearance of the site and complete tissue integration (Stain, hematoxylin and eosin; original magnification, 5). (D) The MACI- grafted area showed good restoration of osteochondral architecture to remove the need for cell injection and cartilage suture, and regenerative tissue generally characterized as hyaline-like car- and simplify the ACI procedure. Our results show that MACI tilage (Stain, hematoxylin and eosin; original magnification, 200). allows manufactured control of cell density for implantation, (E) The hyaline-like cartilage regeneration was also rich in pro- and enables the regeneration of hyaline-like cartilage. Al- teoglycan (Stain, Alcian blue; original magnification, 200). though these results are promising, the patient cohort was too small to confirm clinical efficacy. Larger sample sizes would be required for histological comparison between MACI and untreated cartilage defects, as well as the cor- to their treated lesion. Characterization of the paradigm of relation of qualitative cartilage repair and cell phenotype/ cartilage regeneration by MACI is crucial to the continual density prior to implantation. However, as in our cohort, evaluation of this technology, but may be an unnecessary and patients rarely do or should consent to further unnecessary risky practice for some patients. Considering the shortage intervention to treated defect sites, making controlled histo- of biopsies from our cohort, we acknowledge the current logical evaluation of regenerative tissue extremely difficult. work as a preliminary indication of clinical efficacy, and In fact, there exists an ethical dilemma as to whether patients highlight the need for future large-scale functional and his- should be approached for biopsy given the possible detriment tological analyses to confirm the present findings. Page 74

ZHENG ET AL.

Table 2. SUMMARY OF HISTOLOGIC FINDINGS OF BIOPSIED CASES

Patient Time of biopsy Histological appearance

1 48 hours Chondrolineage cells mixed with ﬁbrin glue 2 21 days Cartilage-like matrix mixed with mesenchymal tissue 3 6 months Hyaline-like cartilage, high cell density, type II collagenþ 4 8 months Hyaline-like cartilage, type II collagenþ 5 12 months I Hyaline-like cartilage, type II collagenþ 6 12 months II Hyaline-like cartilage, type II collagenþ 7 12 months III Mix of hyaline-like and ﬁbrocartilage 8 12 months IV Fibrocartilage 9 18 months Hyaline-like cartilage, type II collagenþ 10 18 months Hyaline-like cartilage, type II collagenþ 11 24 months Hyaline-like cartilage, type II collagenþ

Although it has been shown that chondrocytes cultured for fibrin sealant is an essential component element in the clinical ACI/MACI incur low apoptosis, they have been documented success of MACI.29,42 to cease production of type II collagen and glycosaminogly- Various studies have reported on histological outcomes can, and display a fibroblast-like phenotype with time.10,31,32 subsequent to the implantation of autologous chondrocytes in Chondrocytes transplanted as a cell suspension after mono- patients.11,12,15,16,21,30 Autologous chondrocyte implantation layer culture, as in periosteal ACI, may theoretically be less using periosteum has consistently reported more than 65% capable of differentiating into the chondrocyte phenotype hyaline-like cartilage regeneration, with the remainder com- necessary to facilitate regeneration tissue resembling native prised of inferior fibrocartilage or a fibro/hyaline cartilage hyaline cartilage cell architecture, a feature not demonstrated hybrid tissue.12,16,21 Richardson et al.showedthatfibrocar- by previous cell-injected ACI.11 Direct inoculation of chon- tilage tissue outcome was always in company with positive drocytes on type I/III collagen scaffold stabilizes the phe- type II collagen, suggesting a hyaline-like matrix production notypic profile of chondrolineage cells. Within this study we within these fibroblast-like cells.21 Briggs et al.reporteda have reported that 80% of implanted patient membranes 51%and57%hyaline-likecartilageregeneration,respectively, contain 80% or greater S-100 positive cells (chondrocytes). in two studies after ACI using collagen scaffold.11,15 The The maintenance of the chondrocytic phenotype is important fibrocartilage seen was also type II collagen positive; how- to the maturation of hyaline-specific matrix in articular de- ever, there was also fibrous tissue repair in one cohort.11 fects and for regenerative success, as the fibrocartilaginous Henderson et al. also showed hyaline-like to hyaline cartilage matrix produced by infiltrating fibroblasts appears unsuitable at 12 months in 69% of patients following periosteal ACI.16 for long-term biomechanical competence of the joint.37–39 In Although periosteal ACI has also produced hyaline-like tis- a limited number of patients we also observed no relationship sue regeneration, most outcome evaluations have been con- between the cell number in the implanted patient constructs ducted beyond 1 year postoperatively, leaving questions and their resultant histological outcomes. Further study is relating to the timeframe of regeneration.11,12,15,16 Consid- required to assess any possible association between these ering the mean defect size (5.32 cm2) treated by MACI in this variables, although it is noteworthy that cell number did not cohort is larger than previously reported ACI studies,11,12,20 considerably differ between patients. Although minor resid- we have demonstrated 75% hyaline-like cartilage regenera- ual scaffold collagen was seen in one patient at 12 months, no tion after 6 months. Our data shows early cartilage formation clinical complication was identified. The biocompatibility of after MACI and evidences histology comparable to previous the collagen scaffold for use in ACI may be superior to periosteum because of the lack of reported postoperative complications.11,14,19,25,30,40 Table 3. COMPARISON OF HYALINE-LIKE The reliability of fibrin sealant in incorporating implanted CARTILAGE REGENERATION BETWEEN INVESTIGATORS chondrocyte has also been previously questioned. Brittberg Group (year) Hyalinelike cartilage et al. reported that chondrocytes could not migrate into fibrin glue.41 Our data has illustrated spherical chondrocytes within Present study 75% 14 injected fibrin sealant only 48 hours after implantation. It is Henderson et al (2003) 69% Briggs et al4 (2003) 57% also noteworthy that no complication was noted with the use 12 of fibrin sealant as an adhesive substance for the seeded ACI- Haddo et al (2004) 51.5% Peterson et al30 (2002) 66% Maix collagen scaffold. In addition to our previously reported Brittberg et al5 (1994) 73% animal data, this study has cemented the authors opinion that Page 75

MATRIX-INDUCED AUTOLOGOUS CHONDROCYTE IMPLANTATION studies with other ACI procedures (Table 3).11,12,15,16,20 This A prospective, randomised comparison of autologous chon- study adds weight to the previously documented effective- drocyte implantation versus mosaicplasty for osteochondral ness of this technique in adolescent to middle-aged patients, defects in the knee. J Bone Joint Surg Br 85, 223, 2003. but interestingly, we have also shown that MACI has the 10. Benya, P.D., and Shaffer, J.D. Dedifferentiated chondrocytes ability to regenerate defects with hyaline-like cartilage within reexpress the differentiated collagen phenotype when cultured an arthritic knee joint.11,12,14–16,20,30,43 in agarose gels. Cell 30, 215, 1982. 11. Briggs, T.W., Mahroof, S., David, L.A., Flannelly, J., Pringle, In summary, our data suggests that type I/III collagen J., and Bayliss, M. Histological evaluation of chondral defects scaffold is capable of maintaining the chondrocyte phenotype after autologous chondrocyte implantation of the knee. J Bone upon seeding, representing a viable scaffold for delivering Joint Surg Br 85, 1077, 2003. autologouschondrocytestochondraldefects.Whilstfibrinsea- 12. Brittberg, M., Lindahl, A., Nilsson, A., Ohlsson, C., Isaksson, lant appears to act as an adhesive substance for the seeded O., and Peterson, L. Treatment of deep cartilage defects in the ACI-Maix collagen scaffold, it also facilitates the maturation knee with autologous chondrocyte transplantation. N Engl J of the implanted chondrocytes into functional cartilage. We Med 331, 889, 1994. also found that MACI produces cartilage-like matrix after 21 13. Brittberg, M., Nilsson, A., Lindahl, A., Ohlsson, C., and Peter- days which matures into hyaline-like cartilage at 6 months son, L. Rabbit articular cartilage defects treated with autologous postoperatively. The histological results of this work are con- cultured chondrocytes. Clin Orthop Relat Res 326, 270, 1996. sistent with previous findings; however, a larger cohort is re- 14. Cherubino, P., Grassi, F.A., Bulgheroni, P., and Ronga, M. Autologous chondrocyte implantation using a bilayer collagen quired to ensure the clinical efficacy of MACI. membrane: a preliminary report. J Orthop Surg (Hong Kong) 11, 105, 2003. ACKNOWLEDGMENTS 15. Haddo, O., Mahroof, S., Higgs, D., et al. The use of chondrogide membrane in autologous chondrocyte implantation. Knee 11, 51, 2004. We would like to thank Dr. Keith Holt from Perth Ortho- 16. Henderson, I.J., Tuy, B., Connell, D., Oakes, B., and Hettwer, paedics and Sports Medicine and Dr. Tony Ganko from Bris- W.H. Prospective clinical study of autologous chondrocyte bane Orthopaedic Specialist Services for kindly providing implantation and correlation with MRI at three and 12 months. patient biopsies for this study. Given the difficulty in obtain- J Bone Joint Surg Br 85, 1060, 2003. ing biopsied cartilage, this contribution has been most help- 17. King, P.J., Bryant, T., and Minas, T. Autologous chondrocyte ful. This study was supported by grant funding from the implantation for chondral defects of the knee: indications and National Health and Medical Research Council (NHMRC). technique. J Knee Surg 15, 177, 2002. 18. Minas, T. Autologous chondrocyte implantation for focal chondral defects of the knee. Clin Orthop Relat Res 391 Suppl, REFERENCES S349, October 2001. 19. Nehrer, S., Breinan, H.A., Ramappa, A., et al. Chondrocyte- 1. Bugbee, W.D., and Convery, F.R. Osteochondral allograft seeded collagen matrices implanted in a chondral defect in a transplantation. Clin Sports Med 18, 67, 1999. canine model. Biomaterials 19, 2313, 1998. 2. Convery, F.R., Meyers, M.H., and Akeson, W.H. Fresh os- 20. Peterson, L., Brittberg, M., Kiviranta, I., Akerlund, E.L., and teochondral allografting of the femoral condyle. Clin Orthop Lindahl, A. Autologous chondrocyte transplantation. Biome- Relat Res 273, 139, 1991. chanics and long-term durability. Am J Sports Med 30, 2, 2002. 3. Hangody, L., and Fules, P. Autologous osteochondral mo- 21. Richardson, J.B., Caterson, B., Evans, E.H., Ashton, B.A., and saicplasty for the treatment of full-thickness defects of weight- Roberts, S. Repair of human articular cartilage after implantation bearing joints: ten years of experimental and clinical experi- of autologous chondrocytes. J Bone Joint Surg Br 81, 1064, 1999. ence. J Bone Joint Surg Am 85-A (Suppl 2), 25, 2003. 22. Driesang, I.M., and Hunziker, E.B. Delamination rates of 4. Jackson, R. Arthroscopic Treatment of Degenrative Arthritis. tissue flaps used in articular cartilage repair. J Orthop Res 18, New York: Raven Press, 1991, pp. 319–323. 909, 2000. 5. Johnson, L. Arthroscopic Abrasion Arthroplasty. New York: 23. Minas, T., and Nehrer, S. Current concepts in the treatment of Raven Press, 1991, pp. 341–360. articular cartilage defects. Orthopedics 20, 525, 1997. 6. Louisia, S., Beaufils, P., Katabi, M., and Robert, H. Transchon- 24. Nehrer, S., Spector, M., and Minas, T. Histologic analysis of dral drilling for osteochondritis dissecans of the medial con- tissue after failed cartilage repair procedures. Clin Orthop dyle of the knee. Knee Surg Sports Traumatol Arthrosc 11, 33, Relat Res 365, 149, 1999. 2003. 25. Ueno, T., Kagawa, T., Mizukawa, N., Nakamura, H., Sugahara, 7. Steadman, J.R., Briggs, K.K., Rodrigo, J.J., Kocher, M.S., T., and Yamamoto, T. Cellular origin of endochondral ossifica- Gill, T.J., and Rodkey, W.G. Outcomes of microfracture for tion from grafted periosteum. Anat Rec 264, 348, 2001. traumatic chondral defects of the knee: average 11-year follow- 26. Gooding, C.R., Bartlett, W., Bentley, G., Skinner, J.A., Car- up. Arthroscopy 19, 477, 2003. rington, R., and Flanagan, A. A prospective, ranomised study 8. Tippet, J. Articular Cartilage Drilling and Osteotomy in Oste- comparing two techniques of autologous chondrocyte im- oarthritis of the Knee. New York: Raven Press, 1991, p. 325. plantation for osteochondral defects in the knee: periosteum 9. Bentley, G., Biant, L.C., Carrington, R.W., Akmal, M., covered versus type I/III collagen covered. Knee 13, 203, Goldberg, A., Williams, A.M., Skinner, J.A., and Pringle, J. 2006. Page 76

ZHENG ET AL.

27. Nehrer, S., Domayer, S., Dorotka, R., Schatz, K., Bindreiter, U., lesions in workers’ compensation patients. Orthopedics 26, and Kotz, R. Three-year clinical outcome after chondrocyte 295, 2003; discussion 300–301. transplantation using a hyaluronan matrix for cartilage repair. 37. Malinin, T., and Ouellette, E.A. Articular cartilage nutrition is Eur J Radiol 57, 3, 2006. mediated by subchondral bone: a long-term autograft study in 28. Marcacci, M., Berruto, M., Brocchetta, D., et al. Articular baboons. Osteoarthritis Cartilage 8, 483, 2000. cartilage engineering with Hyalograft C: 3-year clinical results. 38. Newman, A.P. Articular cartilage repair. Am J Sports Med Clin Orthop Relat Res 435, 96, 2005. 26, 309, 1998. 29. Willers, C., Chen, J., Wood, D., Xu, J., and Zheng, M.H. 39. Prakash, D., and Learmonth, D. Natural progression of osteo- Autologous chondrocyte implantation with collagen bioscaf- chondral defect in the femoral condyle. Knee 9, 7, 2002. fold for the treatment of osteochondral defects in rabbits. 40. Russlies, M., Behrens, P., Wunsch, L., Gille, J., and Ehlers, E.M. Tissue Eng 11, 1065, 2005. A cell-seeded biocomposite for cartilage repair. Ann Anat 184, 30. Bartlett, W., Skinner, J.A., Gooding, C.R., et al. Autologous 317, 2002. chondrocyte implantation versus matrix-induced autologous 41. Brittberg, M., Sjogren-Jansson, E., Lindahl, A., and Peterson, chondrocyte implantation for osteochondral defects of the knee: L. Inﬂuence of ﬁbrin sealant (Tisseel) on osteochondral defect a prospective, randomised study. J Bone Joint Surg Br 87, repair in the rabbit knee. Biomaterials 18, 235, 1997. 640, 2005. 42. Kirilak, Y., Pavlos, N.J., Willers, C.R., et al. Fibrin sealant 31. Zheng, M.H., King, E., Kirilak, Y., et al. Molecular charact- promotes migration and proliferation of human articular chon- erisation of chondrocytes in autologous chondrocyte implanta- drocytes: possible involvement of thrombin and protease- tion. Int J Mol Med 13, 623, 2004. activated receptors. Int J Mol Med 17, 551, 2006. 32. Gigante, A., Bevilacqua, C., Ricevuto, A., Mattioli-Belmonte, M., 43. Micheli, L.J., Moseley, J.B., Anderson, A.F., et al. Articular and Greco, F. Membrane-seeded autologous chondrocytes: cell cartilage defects of the distal femur in children and adoles- viability and characterization at surgery. Knee Surg Sports Trau- cents: treatment with autologous chondrocyte implantation. matol Arthrosc, 15, 88, 2007. J Pediatr Orthop 26, 455, 2006. 33. Outerbridge, R.E. The etiology of chondromalacia patellae. J Bone Joint Surg Br 43-B, 752, 1961. 34. Hambly, K., Bobic, V., Wondrasch, B., Van Assche, D., and Address reprint requests to: Marlovits, S. Autologous chondrocyte implantation postoper- Ming-Hao Zheng, Ph.D., D.M., FRCPath ative care and rehabilitation: science and practice. Am J Sports Unit of Orthopaedics Med 34, 1020, 2006. School of Pathology and Surgery 35. Huang, L., Xu, J., Wood, D.J., and Zheng, M.H. Gene expres- University of Western Australia sion of osteoprotegerin ligand, osteoprotegerin, and receptor 2nd Floor M-block, QEII Medical Centre activator of NF-kappaB in giant cell tumor of bone: possible in- Nedlands 6009 volvement in tumor cell-induced osteoclast-like cell formation. Australia Am J Pathol 156, 761, 2000. 36. Yates, J.W., Jr. The effectiveness of autologous chondrocyte implantation for treatment of full-thickness articular cartilage E-mail: [email protected]

Chapter 6

Clinical MACI: Revised & Failed Histology

Thesis publication #5: Willers C, Stoffel K, Zheng MH. Histological assessment of revised and clinically failed matrix -induced autologous chondrocyte implantation. Provisionally accepted into Osteoarthritis and Cartilage, August 2008.

Page 77

STATEMENTS OF AUTHORSHIP CONTRIBUTION FROM CO-AUTHORS

Histological Assessment of Revised and Clinically Failed Matrix-induced Autologous Chondrocyte Implantation.

Willers C, Stoffel K, Zheng MH.

Provisionally accepted into Osteoarthritis and Cartilage, August 2008.

Willers C (PhD Candidate) Major contribution to the planning, execution, analysis, and interpretation of all research. Major contribution to writing of the manuscript.

Signature of Principal Author: ……………………………… Date: .………………

Stoffel K (Research Collaborator) Moderate contribution to the execution of research.

Signature of Co-Author: ……………………..……………… Date: …………………

Zheng MH (Supervisor) Minor contribution to the execution of the research.

Signature of Co-Author: ……………………..……………… Date: …………………

Page 78

STATEMENTS OF PERMISSION FROM CO-AUTHORS

I, Karl Stoffel give permission to Craig Willers, principal author of the paper: Histological Assessment of Revised and Clinically Failed Matrix-induced Autologous Chondrocyte Implantation, provisionally accepted into Osteoarthritis and Cartilage (2008), to include this paper as Chapter 6 of his PhD thesis entitled Matrix- induced Autologous Chondrocyte Implantation for Articular Cartilage Injury: Biology, Histology, and Clinical Outcomes.

Signature: …………………………………. Date: ………………………………….

I, Ming-Hao Zheng give permission to Craig Willers, principal author of the paper: Histological Assessment of Revised and Clinically Failed Matrix-induced Autologous Chondrocyte Implantation, provisionally accepted into Osteoarthritis and Cartilage (2008), to include this paper as Chapter 6 of his PhD thesis entitled Matrix- induced Autologous Chondrocyte Implantation for Articular Cartilage Injury: Biology, Histology, and Clinical Outcomes.

Signature: …………………………………. Date: ………………………………….

Page 79

FULL TITLE: Histological Assessment of Revised and Clinically Failed Matrix- induced Autologous Chondrocyte Implantation (MACI).

SHORT RUNNING TITLE: Revised and Clinically Failed MACI.

C. Willers1, M(Med)Sc; K. Stoffel2, MD PhD; M.H. Zheng1, MD PhD FRCPath

1 Centre for Orthopaedic Research, School of Surgery, University of Western Australia,

QE2 Medical Centre, Nedlands 6009, Western Australia, Australia.

2 Fremantle Orthopaedic Unit, The University of Western Australia, Fremantle Hospital,

Alma Street, Fremantle, Western Australia 6160, Australia

Dr. Craig Willers: TELEPHONE: 08 93463213, FAX: 08 93463210,

EMAIL: [email protected]

Dr. Karl Stoffel: TELEPHONE: 08 94313863, FAX: 08 94312701

EMAIL: [email protected]

Prof. Ming-Hao Zheng: TELEPHONE: 08 93463213, FAX: 08 93463210,

EMAIL: [email protected]

CORRESPONDENCE:

Professor Ming-Hao Zheng MD PhD FRCPath

Centre for Orthopaedic Research, School of Surgery, University of Western Australia,

2nd Floor M-block QEII Medical Centre, Nedlands 6009, Australia

Phone: 08 93463213; Fax: 08 93463210; E-mail: [email protected]

Page 80

ABSTRACT

Objective: To investigate the histological and immunohistochemical characteristics of revised and failed MACI repair tissues.

Methods: We examined the matrix profiles of repair biopsies taken from revised and clinically failed MACI cases by semi-quantitative immunohistochemical study using antibodies specific to aggrecan, collagens I, II, III, VI, and IX, Sox-9, Ki-67 and MMP-

13. We also stiffness tested an intact clinically failed repair site.

Results: Histologically, the majority of these biopsies (n=39) were hyaline-like (HLC) and fibrocartilage (FC) in both the revised (30% and 38% respectively) and failed (34% and 22% respectively) cases. Compositionally, more revised cases were positive for aggrecan, collagens VI and IX, and Ki67 compared to failed cases, but not quantitatively different (P>0.05). More HLC biopsies were positive for aggrecan and collagen II (compared to the FC group), with diffuse and often colocalized matrix distribution. The majority of HLC biopsies stained positive for Sox-9, whereas FC cases were negative. Most (75%) FC biopsies were positive for Ki-67, compared to the HLC group with 25%. MMP-13 was negative in all biopsies. Qualitatively, reduced collagen

II and IX, and increased Ki67 production was noted in FC biopsies (P<0.05). An intact repair site showed FC with 30% greater stiffness in the inferior portion compared to the superior, with an associated proteoglycan content increase.

Conclusions: Revised and failed biopsies display predominantly hyaline-like and fibrocartilage in repair type, are histologically dissimilar to healthy cartilage, but do not differ in composition. Hyaline-like repairs show lower proliferation but improved matrix to fibrocartilage repairs. Our study furthers knowledge into failed and revised cartilage repair by MACI.

Keywords: Matrix-induced autologous chondrocyte implantation; cartilage repair; revised; failed; hyaline-like; fibrocartilage; immunohistochemistry

Page 81

INTRODUCTION

The treatment of articular cartilage injury is troublesome, due mainly to the functional complexities of the tissue, and its inability to contribute to self-regeneration following insult. Indeed, osteochondral autografting (or mosaicplasty) is often technically complicated by attempting to maintain surface continuity, donor/recipient matrix orientation at the graft site, and by inferior fibrous grouting between transplanted cylinders relating poor graft integration 1-5. Similarly, microfracture has been shown to fill chondral defects with predominantly fibrous or fibrocartilage repair, a biomechanically inferior tissue that deteriorates in the joint with time 6,7. It has been postulated that this inferior repair tissue is related to the phenotypic plasticity of the infiltrating marrow cells. Moreover, cells isolated from microfracture repair tissue, compared to autologous chondrocytes, have been shown to lack type II collagen and express osteocalcin, a marker of bone formation 8,9.

As a consequence of these limitations, autologous chondrocyte therapies have become increasingly popular due to their provision of a more stable cell source. But whilst the clinical and histological features of successful autologous chondrocyte implantation

(ACI) have been widely reported, as with the mechanisms driving good outcomes, the repair characteristics of revised and clinically failed grafts are not well understood 10-18.

LaPrade recently illustrated that failed ACI is predominantly composed of fibrous tissue and fibrocartilage with variable positivity for both collagen I and II 10. Although, this study looked at osteochondritis dissecans lesions, and examined periosteal ACI, not the new generation of ACI - matrix-induced ACI (MACI) 10. Notably, histological differences in repair would be expected between these two techniques as the former initiates repair with a cell solution capped with a living tissue (periosteum), whereas the latter initiates repair with a cell-scaffold construct adhered to the subchondral plate 17,18.

Page 82

Hence, investigation into the histological features of revised and clinically failed MACI cases may shed light on the treatment-specific reason for their poor outcome.

The objective of this study was to examine the histological and immunohistochemical characteristics of revised and clinically failed articular cartilage repair sites biopsied following treatment by MACI. We hypothesize that these biopsies will be similar in composition, but different to that known for healthy cartilage.

METHODS

Patient Biopsies

Over 2000 patients have received MACI treatment between 2000 and 2006 in Australia.

The Centre for Orthopaedic Research, University of Western Australia (UWA) is the referred centre for the examination of cartilage tissue in cases requiring revision arthroscopy or cases where the graft has clinically failed and total joint replacement has ensued. Accordingly, a retrospective collection of samples for thirty-nine patients with revised or clinically failed MACI was examined. Of these, all were histomorphologically characterized but only 12 were classified as suitable for immunohistochemical analysis based on the presence of complete repair infill and tissue integrity for the rigors of antigen retrieval and washing. Revised tissues were all swipe cartilage biopsies taken as part of the chondroplasty of symptomatic tissue hypertrophy at the repair site, while failed cases were obtained as osteochondral samples as part of total joint replacement surgery. The location of arthroscopic biopsy removal was recorded in detailed surgical notes, while the treatment site in the osteochondral specimens was examined arthroscopically to insure that only repair tissue was removed for assessment. All patients consented to having their treatment and the use of their

Page 83 biopsy for research, and ethical approval was given by the Human Research Ethics

Committee of University of Western Australia. Patient age ranged from 18 to 54 years

(mean 39.6±9.7) at the time of biopsy. Gender distribution was 17 men to 22 women.

Defects ranged from 1 to 7.5 cm2 (mean 3.6 cm2±3.9) in area. The majority of cases were located on the medial femoral condyle (n=30), with the lateral femoral condyle

(n=3), patella (n=2), talar dome (n=2), and hip (femoral head; n=2) completing the group.

Histology

Routine histology was used primarily to confirm repair tissue type. Biopsies were placed into 4% paraformaldehyde immediately after surgery. The tissues were then decalcified with 10% formic acid, dehydrated by a graded series of alcohol and xylene washes, and paraffin embedded (blocks were prepared for processing by transverse dissection of the defect centre). Sections were then cut to 5μm and stained with haematoxylin and eosin (H&E) or Alcian Blue (AB), before being assessed under light microscopy. Repair tissues were characterized as hyaline-like cartilage (HLC), fibrocartilage (FC), or fibrous tissue (FT) on the basis of their cellularity (density, shape, and lacunae) and matrix appearance (H&E and AB) in accordance with our previous work 17. Nearly all cartilage repairs are mixed tissues upon histological analysis. For this reason, over 80% section area was used to characterize as HLC, FC, or

FT, with smaller percentages termed mixed tissue. Outcomes were blindly assessed by the authors, including a qualified histopathologist (MHZ).

Immunohistochemistry

Following overnight incubation (30°C) on silanated slides, sections were deparaffinised with xylene, rehydrated with decreasing ethanol solutions and rinsed for 10 minutes in

Page 84 distilled water. Specific antibody staining conditions are detailed in Table 1. Sections were boiled in a DAKO pressure cooker at 100ºC for 2 minutes in pH6 citrate buffer, or trypsin treated (Ki67 only) for 30 minutes at RT for antigen retrieval. Sections were then rinsed in 3% H2O2 in methanol for 15 minutes to block endogenous peroxidase activity. Non-specific staining was blocked by rinsing sections in 10% FBS in 0.1% triton X100 in TBS for 30 minutes, before incubation with primary monoclonal anti- human antibodies diluted in 0.1% triton X100/ 1% BSA in TBS for a specified time at

RT. After washing 3 times for 5 minutes in 1X TBS, sections were incubated in

EnVision+ goat anti-mouse IgG peroxidase secondary antibody system at RT for 30 minutes. Sections were then washed 3 times for 5 minutes in TBS, before incubation in their associated chromogen detection system (DAB; BCIP/NBT for Ki67 only) at RT for a specified time. Sections were washed in distilled water, counterstained shortly with haematoxylin (except Ki67 – not counterstained), washed, stained shortly with Scott’s tap water, washed in distilled water, then dehydrated, cleared and mounted. Positive

(healthy cartilage: agg, coll II/III/VI/IX, sox-9; testis: sox-9; colon: ki-67; placenta and giant-cell tumor: MMP-13) and negative controls were included in every run. Slides were then analysed for signal intensity using Aperio Scanscope (Vista, CA) hardware and software. Color deconvolution algorithm using default manufacturer threshold settings calculated the percentage negative, weak positive, medium positive, and strong positive antibody staining for all matrix protein antibodies. The nuclear factors, Ki67 and Sox9, were analyzed by counting five random areas (500µm x 500µm) in each slide.

Tissue Indentology

A complete medial femoral condyle fibrocartilage repair (post-MACI) site was retrieved and immediately frozen at -80ºC until testing. The sample was selected because it

Page 85 showed complete infill and good macroscopic appearance and surface continuity for compression testing. The case was undergoing knee replacement because of concomitant pathology not related to the repair site. The condyle was thawed in a bath of Ringer’s solution for approximately 45 min and dissected into subregions to test for intra-defect stiffness variability. Four subregions were tested: M1 and M2 (superior and inferior aspects of repair site respectively), and C1 and C2 (healthy cartilage immediately adjacent to the repair site). All samples were embedded in bone cement

(CMW1 Radiopaque, Depuy) and positioned appropriately for surface indentation

(perpendicular to the indentation tip). The lateral edges of the specimens were not in contact with the cement to allow deformation of the cartilage to all sides. Thus, the internal load state was not influenced by the surrounding specimen holders. Specimens were moisture-controlled with phosphate buffered saline. The specimen and its immediately adjacent osteochondral tissue was indentation tested on a material testing machine (Zwick Z010, Zwick Inc., Ulm, Germany). After a preload of 1 N, creep tests were performed at a rate of displacement of 0.5 mm/min up to a deformation

/indentation of 0.3mm. Loads were applied to the cartilage using a 1.2mm diameter K- wire. The deformation was measured with the linear variable differential transformer on the Zwick Machine with a resolution of 0.06 um. The load needed for this deformation was recorded. The testing machine was controlled by a standard IBM PC with Zwick software (test-expert-Software; Zwick inc., Ulm, Germany). The measured data were directly transferred and stored in ASCII files to the same computer. Each region was stiffness tested six times for reliability. Stiffness values were expressed a mean of the six values recorded ± SEM. All tested samples were then histologically processed to confirm tissue type.

Page 86

Statistics

All data was stored on Excel spreadsheets, and graphed using Excel (Microsoft, Seattle,

USA) and Photoshop (Adobe, San Jose, USA) software. Staining intensities were calculated by Aperio ImageScope software (v7.1.100.1248). The student’s t-tests was used to test for significance (P<0.05) between immunohistochemical outcomes. The

Kruskal-Wallis test was used to test significance (P = 0.01) between the results of the indentation tests for different repair localization.

RESULTS

Histology and Immunohistochemistry of Repaired Cartilage

The majority of biopsies were characterized by HLC and FC repair in both the revised

(30% and 38% respectively) and failed (34% and 22% respectively) cases (Table 2).

Lesions on the medial femoral condyle (77% of cases), as with overall, composed chiefly of HLC (40% of cases) and FC (30% of cases), although no HLC repair was recorded in any of the non-femoral sites (Table 3). Revised cases showed a greater frequency of positivity for aggrecan, collagens VI and IX, and Ki67 compared to failed cases (Table 4). Notably, the pericellular-specific distribution of collagen VI was lost in all repairs. However, as seen in Figure 1, of the nine chondral markers examined, we found no significance difference (P<0.05) in percentage positivity between revised and clinically failed biopsies (for weak, medium, strong, and overall percentage positivity).

It is also noteworthy that no residual fibrin sealant (a component of the MACI technique) was observed in any samples.

To further characterize hyaline-like and fibrocartilage repair we applied the same antibody array for analysis (Figures 2, 3, & 4). HLC was characterized (H&E) by a

Page 87 sparse but seemingly uniform distribution of mature chondrocytes within their lacunae, surrounded by hyaline-like matrix. The HLC group showed positive aggrecan and collagen II expression (50% and 75% of cases respectively) with diffuse, and often colocalized matrix distribution in the majority of cases. In relation, 87% of cases were positive for Sox-9, localized to the nuclear region. The HLC group was also positive for the matrix collagens III, VI and IX (75%, 75%, and 62% of cases respectively).

Interestingly, most collagen VI localization was pericellular. Collagen I was detected in

25% of cases, but was only weakly positive. HLC was negative for MMP-13, and showed low (25% of cases) positivity for the proliferative marker Ki-67.

FC was characterized (H&E) by an irregular distribution of predominantly spindle- shaped chondrocytes within a dense irregular fibrous matrix. In contrast, the FC group was negative for the hyaline-specific matrix proteins aggrecan and collagen II (12% and

25% of cases respectively) in most cases, and accordingly was negative for Sox-9 expression. Figure 3 shows representative staining. The FC group was also positive for the matrix collagens III, VI and IX (75%, 50%, and 62% of cases respectively) in the majority of cases, with similar matrix localization. Collagen VI localization was diffuse in localization. Surprisingly, except one case, the FC group was negative for collagen I, a common component in fibrous repair. The FC group was also negative for MMP-13.

Lastly, 75% of the FC cases were positive for Ki-67, suggesting that FC has an increased proliferative profile to HLC.

We evidenced a significant (P<0.05) reduction in collagen II expression in the FC group compared to both HLC and HLC/FC mix groups (Figure 4). With the exception of one case, all biopsies positive for Sox-9 where positive for either aggrecan or collagen II expression. Moreover, we observed a significant reduction in collagen IX in the FC

Page 88 group, a collagen known to provide cross-linking strength to the collagen II network.

We also showed a significant (P<0.05) increase in Ki67 in the FC group, suggesting a greater comparative proliferation capacity. There was no observable relationship between patient demographics (age, defect area, or cell number implanted) and biopsy profile.

Stiffness of Repaired Cartilage

We also acquired an entire post-MACI medial femoral condyle cartilage repair site as discarded during total knee replacement, to test the biomechanical stiffness of the FC repair. Indentology of the superior and inferior portions of the MACI graft revealed tissue with 55% and 85% (respectively) the stiffness of the adjacent cartilage (Figure 5).

Notably, histochemical (AB) follow-up noted that the superior portion was characterized by fibrocartilage with negligible proteoglycan, whereas the inferior portion had a similar proteoglycan staining profile to the native adjacent cartilage. The increased stiffness seen in the inferior portion was most likely due to its greater proteoglycan content, a phenomenon possibly explained by greater compressive stimuli in that region of the knee.

DISCUSSION

As with the biological mechanisms underlying good to excellent clinical outcome after

MACI, little is known about the factors which lead to revision and clinical failure following this technique. Hence, profiling the repair type and composition of these tissues may allow us to better understand their functional behaviour and possible relationship to clinical outcome. To this end, we have histologically typed thirty-nine revised and clinically failed patient biopsies, and both qualitatively and quantitatively

Page 89 examined their composition using immunohistochemistry. Importantly, revised and clinically failed biopsies were not significantly different in composition, and the majority of cases were hyaline-like and fibrocartilage. More HLC biopsies were positive for aggrecan, collagen II, and sox9 expression, compared to the FC group. Whereas more FC biopsies displayed Ki-67 production compared to the HLC group, suggesting a higher proliferative capacity. Qualitatively, reduced collagen II and IX, and increased

Ki67 production was observed in FC biopsies. While indentology analysis revealed significant biomechanical variation within the failed repair site, likely due to proteoglycan variation. Our findings provide a greater insight into the biological properties of MACI revision and clinical failure.

Similar to previous studies looking at the histology of failed ACI, we found that a high percentage of our revised and failed repair sites were composed of fibrocartilage; however, a high percentage of our biopsies were also hyaline-like repair 10,11. This disparity in findings may be accounted for by the different surgical techniques used

(periosteal ACI versus MACI). Additionally, our data suggests that revised cartilage repair is similar in composition to clinically failed MACI repairs. However, conclusions drawn from this finding should be approached with scrutiny. Firstly, biopsies obtained from revised cases are generally slice biopsies from the surface of the repair site. Hence it is unclear whether this hypertrophic surface represents a tissue composition indicative of the repair as a whole, or only the upper portion of the repair. Secondly, patients classified as clinically failed may be completely or partially a consequence of other pathologic changes within the joint and not directly related to the MACI treatment. In other words, clinically failed cases may contain a biologically and functionally adequate repair site. Tighter control on patient inclusion criteria may help to confine the cause of failure to the repair site for more robust discussion.

Page 90

The present study has demonstrated greater collagen II and aggrecan expression in hyaline-like cartilage repairs than in fibrocartilage repairs 14,17,19-22. Well understood to be key matrix components for healthy articular cartilage function, aggrecan and collagen II expression is crucial to the biomechanical strength of cartilage repair tissue, and their reduced expression in fibrocartilage explains the weaker comparative strength of the tissue. This idea was consistent with our investigation of a complete MACI repair site characterized by fibrocartilage, whereby reduced aggrecan staining translated a weaker (55%) stiffness under axial compression. Furthermore, we have demonstrated that aggrecan expression varied widely, both between biopsied samples and within a single repair site. Indeed, the superior portion of our complete MACI repair site displayed negligible aggrecan, compared to the inferior portion of the repair which stained much stronger for aggrecan and consequently had higher normalized stiffness.

Although only seen in one sample, this difference in aggrecan distribution and subsequent stiffness may be attributed to differential patellofemoral force in the joint

23,24. Following on this line, the anatomical location of our biopsied sites and the associated variations in patellofemoral load distribution across our cohort may also account for some of the observed differences in the protein array evaluated herein.

Sox-9 is known to play a key role in limb chondrogenesis. This important transcription factor has been shown to increase expression of the chondrocyte-specific genes collagen

II/ IX/ XI and aggrecan, and it’s mutation can cause lethal skeletal malformation 25-28.

Grigolo et al have previously noted the presence of Sox-9 in cartilage repair tissue two years after periosteal ACI, although the study size was small and appeared to be dominated by variable biopsy tissue 20. Our study has shown Sox-9 positivity in 7 of 8 hyaline-like biopsies, but only 1 of 8 fibrocartilage samples. This finding highlights the closer biological relationship of HLC to healthy hyaline articular cartilage, and hence

Page 91 affirms the term ‘hyaline-like’. Whilst the regulatory elements involved in Sox-9 driven chondrogenesis and the regulators of Sox-9 expression are still to be clarified, it would appear that these factors are reactivated during repair classified as hyalinelike, but not during fibrocartilage repair. Furthermore, the predominance of Sox-9 expression in the

HLC cases supports the use of the transcription factor as a qualitative marker for assessing repair outcomes following MACI.

Greater proliferation was noted in the FC biopsies in the current study, with threes times as many biopsies positive for Ki67 compared to the HLC group. Accordingly, FC repairs appear to grow into the defect void faster than HLC repairs, yet exhibit less hyaline-specific matrix expression, particularly collagen II. This may be in part explained by the differential collagen VI (known to translate biomechanical stimuli into matrix synthesis) localization between FC and HLC cartilage repair. Whilst the frequency of collagen expression was similar (except type II) across both groups in the current study, the localization of collagen VI was isolated to the pericellular region in

HLC repairs (as in the healthy cartilage), but stained diffusely in the FC cases. Whilst the exact mechanism and importance of collagen VI is unclear, its pericellular distribution (typical chondron structure) has been found to present only in hyaline or hyaline-like cartilage 29. Moreover, and consistent with our findings, the chondron structure is known to localize to regions of extracellular matrix rich in collagen II 29.

Hence, although FC repair may proliferate faster, its matrix synthesis and subsequent biomechanical strength may be inhibited by its lack of hyaline-specific histology, possibly due to altered biomechanical stimuli transduction.

Page 92

Although this study used a valid selection of matrix proteins and nuclear factors to partially characterize differences between hyaline-like cartilage and fibrocartilage repair, a notable limitation of the study is that, relative to the many possibilities, only nine antibodies were used. Nonetheless, given the appreciable expense of large scale antibody array analysis, we have carefully chosen this selection due to their crucial involvement in chondrogenesis, tissue differentiation, and cartilage function. Secondly, although akin to similar studies, our cohort was reasonably small at twelve cases.

Unfortunately, this low sample size is an inherent limitation to any studies assessing cartilage repair biopsies, as patients (and surgeons) are understandably reluctant to undertake further invasive intervention, especially in cases with successful repair.

Ideally, a large-scale study (>50 cases) using tissue microarray technology and a much larger antibody selection would be the next step forward in determining a more detailed comparative analysis of these two cartilage repair types. In general, it should also be noted that the histological assessment of biopsied repair tissue may not represent the greater tissue in all cases. For example, shave biopsies only represent the surface of the repair tissue, a region frequently composed of more fibrous matrix than the underlying tissue. Similarly, core biopsies can be taken from regions not representative of the majority of the repair, therefore falsely characterizing the holistic tissue outcome.

Asides from multi-site biopsy, the most feasible method of bypassing this limitation is the development of a non-invasive imaging system capable of detailed characterization of the entire repair site. Whilst contrast-enhanced magnetic resonance imaging (MRI) offers some indication of the quality of cartilage repair, the information offered by this technology requires substantial improvement before it can replace conventional histological assessment.

Page 93

In summary, we found that revised and clinically failed biopsies are both mainly hyaline-like and fibrocartilage in repair type, but do not significantly differ in composition. Whilst in terms of repair type, fibrocartilage biopsies possess more Ki-67 production compared to hyaline-like, suggesting a higher proliferative capacity.

Qualitatively, fibrocartilage biopsies show reduced collagen II and IX, and increased

Ki67 production. While biomechanical testing revealed significant variation within the failed repair site, corresponding to proteoglycan variation. Our findings provide a greater insight into the biological properties of MACI revision and clinical failure, while suggesting further investment into improving the efficacy of this technique may be needed.

ACKNOWLEDGEMENTS

We would like to thank Dr. Ray Crowe and Dr. David Wood for kindly providing patient biopsies for this study. Given the difficulty in obtaining biopsied cartilage, this contribution has been most helpful. This study was supported by grant funding from the

National Health and Medical Research Council (NHMRC).

Page 94

Figure 1: Repair tissue composition for revised and clinically failed cases. Whilst small differences were observed between the composition of revised and failed cases, none were found to be significant (P>0.05). BARS (top two rows): Light grey = weak positive; Dark grey = medium positive; Black = strong positive; Dotted white = total positivity. BARS (MMP13, Sox9, Ki67): Black = total nuclei positive. Sample size: n=12 for each antibody. Error bars indicate standard error (SEM).

Page 95

Figure 2: Representative histology and immunohistochemical images of the hyaline-like cartilage (HLC) cohort. HLC was characterized (H&E) by a sparse but seemingly uniform distribution of mature chondrocytes within their lacunae, surrounded by hyaline- like matrix. Aggrecan, collagen II, collagen III, collagen VI, collagen IX, Sox-9 were generally positive. Aggrecan and all collagens were diffusely stained within the matrix of the positive tissues. Notably, collagen VI was not localised to the pericellular region of the positive biopsies. While the chondrogenic transcription factor Sox-9 was isolated to the nucleus. Collagen I stained weakly in only a few cases. Ki-67 and MMP-13 were negative in general. Magnifications – 50X: coll I; 100X: agg, coll II, coll III, coll IX, Ki- 67 and MMP-13; 200X: coll VI and Sox-9. Page 96

Figure 3: Representative histology and immunohistochemical images of the fibrocartilage (FC) repair cohort. FC was characterized (H&E) by an irregular distribution of predominantly spindle-shaped chondrocytes within a dense irregular fibrous matrix. Collagen III, collagen VI, collagen IX, and Ki-67 were generally positive. Collagens III, VI, and IX were diffusely stained within the matrix of the positive tissues, while the proliferative marker Ki-67 was isolated to the nucleus (no counterstain). Whereas aggrecan, collagen I, collagen II, Sox-9 and MMP-13 were negative in general. Magnifications - 100X: agg, coll I, coll II, coll III, coll VI, coll IX and MMP-13; 200X: Sox-9 and Ki-67.

Page 97

Figure 4: Tissue composition for hyaline-like, fibrocartilage, and mixed repair cases. Collagen II production was significantly (P<0.05) increased in the both the HLC and mixed repairs compared to FC repairs. Collagen IX production was significantly (P<0.05) increased in HLC repairs compared to FC repairs. While a significant (P<0.05) increase in the proliferative marker Ki67 was observed in the FC repairs. BARS (top two rows): Light grey = weak positive; Dark grey = medium positive; Black = strong positive; Dotted white = total positivity. BARS (MMP13, Sox9, Ki67): Black = total nuclei positive. Sample size: n=12 for each antibody. Error bars indicate standard error (SEM).

Page 98

Figure 5: Stiffness testing and associated histology of a complete cartilage repair site on the medial femoral condyle. (A) The superior portion of the repair site (M1) displayed a maximum stiffness of 4.32N, which was 55% that of the control adjacent tissue (C1) at 7.73N. (B) This reduced stiffness correlated to a lack of proteoglycan in M1 compared to C1. (A) In contrast, the inferior portion of the repair site (M2) displayed a maximal stiffness of 6.64N, which was 85% that of the adjacent control tissue (C2) at 7.81N. (B) This improved stiffness appears to be correlated to an observable increase in proteoglycan content in M2 compared to M1. All images shown are Alcian Blue staining. Magnification – 50X for all images.

Page 99

Table 1: Antibody list with associated antigen retrieval and dilution factor used.

Page 100

Table 2: Repair type distribution for revised and clinically failed repair biopsies.

Page 101

Table 3: Repair type distribution and clinical status by anatomical location of defect.

Table 4: Demographics, implant details, and repair compositions for revised and clinically failed biopsies.

Page 102 Page 103

References

1. Huang, F.S., Simonian, P.T., Norman, A.G. & Clark, J.M. Effects of small

incongruities in a sheep model of osteochondral autografting. Am J Sports Med

32, 1842-1848 (2004).

2. Kordas, G., Szabo, J.S. & Hangody, L. The effect of drill-hole length on the

primary stability of osteochondral grafts in mosaicplasty. Orthopedics 28, 401-

404 (2005).

3. Whiteside, R.A., Jakob, R.P., Wyss, U.P. & Mainil-Varlet, P. Impact loading of

articular cartilage during transplantation of osteochondral autograft. J Bone Joint

Surg Br 87, 1285-1291 (2005).

4. Kordas, G., Szabo, J.S. & Hangody, L. Primary stability of osteochondral grafts

used in mosaicplasty. Arthroscopy 22, 414-421 (2006).

5. Bentley, G., et al. A prospective, randomised comparison of autologous

chondrocyte implantation versus mosaicplasty for osteochondral defects in the

knee. J Bone Joint Surg Br 85, 223-230 (2003).

6. Frisbie, D.D., et al. Early events in cartilage repair after subchondral bone

microfracture. Clin Orthop Relat Res, 215-227 (2003).

7. Kreuz, P.C., et al. Is microfracture of chondral defects in the knee associated

with different results in patients aged 40 years or younger? Arthroscopy 22,

1180-1186 (2006).

8. Robinson, D. & Nevo, Z. Articular Cartilage Chondrocytes are more

Advantageous for Generating Hyaline-like Cartilage than Mesenchymal Cells

Isolated from Microfracture Repairs. Cell Tissue Bank 2, 23-30 (2001).

9. Dorotka, R., et al. Repair of articular cartilage defects treated by microfracture

and a three-dimensional collagen matrix. Biomaterials 26, 3617-3629 (2005).

Page 104

10. LaPrade, R.F., Bursch, L.S., Olson, E.J., Havlas, V. & Carlson, C.S. Histologic

and immunohistochemical characteristics of failed articular cartilage resurfacing

procedures for osteochondritis of the knee: a case series. Am J Sports Med 36,

360-368 (2008).

11. Nehrer, S., Spector, M. & Minas, T. Histologic analysis of tissue after failed

cartilage repair procedures. Clin Orthop Relat Res, 149-162 (1999).

12. Bartlett, W., et al. Autologous chondrocyte implantation versus matrix-induced

autologous chondrocyte implantation for osteochondral defects of the knee: a

prospective, randomised study. J Bone Joint Surg Br 87, 640-645 (2005).

13. Haddo, O., et al. The use of chondrogide membrane in autologous chondrocyte

implantation. Knee 11, 51-55 (2004).

14. Peterson, L., Brittberg, M., Kiviranta, I., Akerlund, E.L. & Lindahl, A.

Autologous chondrocyte transplantation. Biomechanics and long-term

durability. Am J Sports Med 30, 2-12 (2002).

15. Cherubino, P., Grassi, F.A., Bulgheroni, P. & Ronga, M. Autologous

chondrocyte implantation using a bilayer collagen membrane: a preliminary

report. J Orthop Surg (Hong Kong) 11, 10-15 (2003).

16. Richardson, J.B., Caterson, B., Evans, E.H., Ashton, B.A. & Roberts, S. Repair

of human articular cartilage after implantation of autologous chondrocytes. J

Bone Joint Surg Br 81, 1064-1068 (1999).

17. Zheng, M.H., et al. Matrix-induced autologous chondrocyte implantation

(MACI): biological and histological assessment. Tissue Eng 13, 737-746 (2007).

18. Brittberg, M., et al. Treatment of deep cartilage defects in the knee with

autologous chondrocyte transplantation. N Engl J Med 331, 889-895 (1994).

Page 105

19. Moriya, T., et al. Evaluation of reparative cartilage after autologous chondrocyte

implantation for osteochondritis dissecans: histology, biochemistry, and MR

imaging. J Orthop Sci 12, 265-273 (2007).

20. Grigolo, B., et al. Molecular and immunohistological characterization of human

cartilage two years following autologous cell transplantation. J Bone Joint Surg

Am 87, 46-57 (2005).

21. Briggs, T.W., et al. Histological evaluation of chondral defects after autologous

chondrocyte implantation of the knee. J Bone Joint Surg Br 85, 1077-1083

(2003).

22. Gobbi, A., et al. Patellofemoral full-thickness chondral defects treated with

Hyalograft-C: a clinical, arthroscopic, and histologic review. Am J Sports Med

34, 1763-1773 (2006).

23. Star, M.J., Kaufman, K.R., Irby, S.E. & Colwell, C.W., Jr. The effects of patellar

thickness on patellofemoral forces after resurfacing. Clin Orthop Relat Res, 279-

284 (1996).

24. Powers, C.M., Chen, Y.J., Scher, I. & Lee, T.Q. The influence of patellofemoral

joint contact geometry on the modeling of three dimensional patellofemoral joint

forces. J Biomech 39, 2783-2791 (2006).

25. Bi, W., Deng, J.M., Zhang, Z., Behringer, R.R. & de Crombrugghe, B. Sox9 is

required for cartilage formation. Nat Genet 22, 85-89 (1999).

26. Kawakami, Y., Rodriguez-Leon, J. & Belmonte, J.C. The role of TGFbetas and

Sox9 during limb chondrogenesis. Curr Opin Cell Biol 18, 723-729 (2006).

27. Sekiya, I., et al. SOX9 enhances aggrecan gene promoter/enhancer activity and

is up-regulated by retinoic acid in a cartilage-derived cell line, TC6. J Biol Chem

275, 10738-10744 (2000).

Page 106

28. Kolettas, E., Muir, H.I., Barrett, J.C. & Hardingham, T.E. Chondrocyte

phenotype and cell survival are regulated by culture conditions and by specific

cytokines through the expression of Sox-9 transcription factor. Rheumatology

(Oxford) 40, 1146-1156 (2001).

29. Fraser, S.A., et al. Localization of Type VI Collagen in Tissue-Engineered

Cartilage on Polymer Scaffolds. Tissue Eng (2006).

Chapter 7

Clinical MACI: Functional and Structural Assessment

Thesis publication #6: Robertson WB, Willers C, Wood DJ, et al. Matrix-induced autologous chondrocyte implantation (MACI)at two years: MRI and functional evaluation. Accepted to The Knee, February 2008.

Page 107

STATEMENTS OF AUTHORSHIP CONTRIBUTION FROM CO-AUTHORS

Matrix-induced autologous chondrocyte implantation (MACI) at two years: MRI and functional evaluation

Robertson WB, Willers C, Wood DJ, Linklater JM, Zheng MH, Ackland TR.

Accepted to The Knee, February 2008.

Robertson WB (Research Collaborator) Moderate contribution to the planning of research. Major contribution to the execution, analysis and interpretation of functional outcome research. Moderate contribution to writing of the manuscript.

Signature of Principal Author: ……………………………… Date: .………………

Willers C (PhD Candidate) Moderate contribution to the planning of research. Major contribution to the interpretation of the research. Major contribution to writing of the manuscript.

Signature of Co-Author: ……………………..……………… Date: …………………

Wood D (Co-Supervisor) Moderate contribution to the planning of the research. Minor contribution to execution of research. Minor contribution to writing of the manuscript.

Signature of Co-Author: ……………………..……………… Date: …………………

Page 108

Linklater J (Research Collaborator) Moderate contribution to the execution of the research (MRI). Moderate contribution to the analysis of research (all MRI data).

Signature of Co-Author: ……………………..……………… Date: …………………

Zheng MH (Supervisor) Moderate contribution to the execution of the research (MRI). Moderate contribution to the analysis of research (all MRI data).

Signature of Co-Author: ……………………..……………… Date: …………………

Ackland T (Research Collaborator) Major contribution to the planning of the research. Minor contribution to the writing of the manuscript.

Signature of Co-Author: ……………………..……………… Date: …………………

Page 109

STATEMENTS OF PERMISSION FROM CO-AUTHORS

I, William Robertson give permission to Craig Willers, second author of the paper: Matrix-induced autologous chondrocyte implantation (MACI) at two years: MRI and functional evaluation, accepted to The Knee (February 2008), to include this paper as Chapter 7 of his PhD thesis entitled Matrix-induced Autologous Chondrocyte Implantation for Articular Cartilage Injury: Biology, Histology, and Clinical Outcomes.

Signature: …………………………………. Date: ………………………………….

I, David Wood give permission to Craig Willers, second author of the paper: Matrix- induced autologous chondrocyte implantation (MACI) at two years: MRI and functional evaluation, accepted to The Knee (February 2008), to include this paper as Chapter 7 of his PhD thesis entitled Matrix-induced Autologous Chondrocyte Implantation for Articular Cartilage Injury: Biology, Histology, and Clinical Outcomes.

Signature: …………………………………. Date: ………………………………….

I, James Linklater give permission to Craig Willers, second author of the paper: Matrix-induced autologous chondrocyte implantation (MACI) at two years: MRI and functional evaluation, accepted to The Knee (February 2008), to include this paper as Chapter 7 of his PhD thesis entitled Matrix-induced Autologous Chondrocyte Implantation for Articular Cartilage Injury: Biology, Histology, and Clinical Outcomes.

Signature: …………………………………. Date: ………………………………….

Page 110

I, Ming-Hao Zheng give permission to Craig Willers, second author of the paper: Matrix-induced autologous chondrocyte implantation (MACI) at two years: MRI and functional evaluation, accepted to The Knee (February 2008), to include this paper as Chapter 7 of his PhD thesis entitled Matrix-induced Autologous Chondrocyte Implantation for Articular Cartilage Injury: Biology, Histology, and Clinical Outcomes.

Signature: …………………………………. Date: ………………………………….

I, Tim Ackland give permission to Craig Willers, second author of the paper: Matrix- induced autologous chondrocyte implantation (MACI) at two years: MRI and functional evaluation, accepted to The Knee (February 2008), to include this paper as Chapter 7 of his PhD thesis entitled Matrix-induced Autologous Chondrocyte Implantation for Articular Cartilage Injury: Biology, Histology, and Clinical Outcomes.

Signature: …………………………………. Date: ………………………………….

Page 111

Matrix-induced autologous chondrocyte implantation (MACI) at two years: MRI and functional evaluation.

W.B. Robertson MSc*†, C Willers M(Med)Sc*, J.M. Linklater FRANZCR◊, D.J. Wood BSc MBBS MS FRCS FRACS*, M.H. Zheng DM PhD FRCPath*, T.R. Ackland PhD FASMF†

* Centre for Orthopaedic Research, School of Surgery and Pathology , University of Western Australia, Crawley, WA 6009 Australia. † School of Human Movement and Exercise Science, University of Western Australia, Crawley, WA 6009 Australia. ◊ North Sydney Orthopaedic and Sports Medicine Centre, Crows Nest, NSW 2065 Australia.

W.B. Robertson: [email protected] C Willers: [email protected] J.M. Linklater: [email protected] D.J. Wood: [email protected] M.H. Zheng: [email protected] T.R. Ackland: [email protected]

Correspondence: Mr William Brett Robertson University of Western Australia 35 Stirling Highway Crawley, WA 6009 AUSTRALIA Fax +61 89 346 6462 Email: [email protected]

Page 112

ABSTRACT

The matrix-induced autologous chondrocyte implantation (MACI) technique addresses problems associated with conventional ACI by replacing the use of periosteum with collagen membrane, whilst also showing comparable clinical outcome to conventional treatments. Herein we present MRI and functional outcomes for 31 MACI patients over a 24 month follow-up period. Following MACI knee surgery and coordinated rehabilitation, functional outcomes were measured using the KOOS score and the six- minute walk test at 3, 6, 12, and 24 months. High resolution MRI scoring was used to describe the quality and quantity of the repair tissue at 3, 12, and 24 months in terms of defect infill, signal intensity, surface contour, structure, border integration, subchondral lamina, subchondral bone, and effusion. Patients demonstrated a significant (P<0.001) improvement in walk distance and all five KOOS subscales from 3 to 24 months after

MACI surgery, with the most substantial gains in the first 12 months. Similarly, patients also demonstrated significant (P<0.001) improved MRI scoring from 3 to 24 months, with post-hoc analysis demonstrating improvement predominantly in the first 12 months, then plateauing thereafter. A 10% incidence of hypertrophic growth following

MACI was observed. Interestingly, we found that MRI score significantly (P<0.01) correlated to all functional outcome parameters, whilst defect size and cell number showed no correlation (P>0.05) to any functional parameters. These data give clinical and radiographical support to MACI as treatment for articular cartilage injury, while suggesting that MRI may be used to predict functional change.

Page 113

INTRODUCTION:

Conventional autologous chondrocyte implantation (ACI) was the first surgical technique to highlight the therapeutic potential of autologous cell therapy in the field of orthopaedics [1,2]. However, the original surgical technique described by Peterson et al, requires the use of a periosteal patch (ACI-P), which has been noted to have numerous postoperative complications in the literature [1,3-8]. Subsequently, the use of a collagen membrane in place of periosteum has been advocated, and related studies have indicated that ACI using a type I/III collagen membrane (ACI-C) produces clinical, radiographical and histological outcomes comparable to ACI-P [9-15]. But importantly, in comparison to ACI-P, the clinical outcomes gained with ACI-C have been accompanied by a reported decrease in the incidence of postoperative complications

[11]. But while ACI-C has been shown to exhibit commendable postoperative outcomes, its surgical technique remains cumbersome. A large surgical incision is required in order to microsuture the membrane to the circumference of the defect - a tedious task that increases the length and technical difficulty of the surgery.

Furthermore, concern remains regarding the uneven distribution of chondrocytes within the fluid suspension, possible leakage of suspension fluid through the graft-cartilage interface, and the creation of microdefects in the native cartilage by the suturing process

[10,14,16].

Accordingly, matrix-induced autologous chondrocyte implantation (MACI) has introduced the concept of direct cell inoculation onto the collagen scaffold for implantation [17]. Since the first introduction of the MACI technique in 1998, more than 4000 patients have been treated globally. In this procedure the chondrocytes are no longer injected under a collagen membrane into a sealed defect compartment; instead they are directly seeded onto the same type I/III collagen but are delivered into the

Page 114 chondral defect as a cell-scaffold construct secured to the defect base by a thin layer of fibrin sealant. Figure 1 outlines the biological paradigm of MACI cartilage regeneration.

The MACI procedure can be performed arthroscopically or through mini-arthrotomy depending upon the defect location [10,18]. Cherubino first supported the MACI technique, publishing improved clinical outcomes, no complications, and MRI- visualized hyaline-like repair at early follow-up [10]. Moreover, Basad has reported significantly better functional outcomes (Lysholm-Gillquist) after MACI compared to microfracture at 24 months [19]. Whilst a recent 5 year follow-up of MACI also reported significant clinical improvement using three different outcome measures, further suggesting that the technique is a suitable cell-based treatment in cartilage repair

[20].

In terms of improving clinical outcome, it has been suggested in the literature that early mobilization after ACI may translate a reduced incidence of postoperative knee stiffness and graft failure [21]. With early mobilization using continuous passive motion (CPM) we have conducted a prospective patient cohort study following MACI treatment for chondral injury to the knee. Specifically, we have sought to assess the progressive functional and structural restoration of the cartilage defects after MACI in the first two years, and examine the relationship between these two outcome parameters.

Page 115

MATERIALS AND METHODS:

Patients

A consecutive series of 31 implantations were performed in 28 patients (18 male; 10 female) between August 2001 and March 2004. Thirty-one implantations survived to a minimum of 24 months, however, one patient was excluded from MRI evaluation due to suffering claustrophobia. The mean age at assessment of the clinical outcomes of MACI for focal chondral defects of the knee was 36.5 years (range: 13-60 years) and mean

BMI was 25.9 (range: 17.2–33.9). The main anatomical site was the medial femoral condyle (55%), with the patella (19%), lateral femoral condyle (13%) and trochlea

(13%) making up the remainder. All subjects suffered from persistent pain associated with full thickness chondral lesions (Outerbridge grade III or IV [22], range: 1.5–9.6 cm2), with no clinical sign of bi- or tri-compartmental osteoarthritis as diagnosed by preoperative MRI and confirmed at arthroscopic biopsy. The main etiology was trauma

(45%). Previous surgical procedures are documented in Table 1. Patients were recruited based on the following inclusion/exclusion criteria. Age: 13–60 years; Defect location: medial or lateral femoral condyle, trochlea, or patella (non-opposing lesions only); Area and depth: < 10cm2, down to stable subchondral bone plate; Aetiology: trauma or osteochondritis dissecans; Joint condition: absence of progressive inflammatory disease or osteoarthritis; Joint stability: absence of full meniscectomy or instability; Abnormal weight-bearing: absence of significant varus/valgus abnormality (>5°), patella maltracking, or obesity (body mass index >35); Sensitivities: no history of gentamycin sensitivity; and Compliance: must consent to surgery and be able and willing to partake in rehabilitation exercises.

Page 116

Surgical Technique and Rehabilitation

The surgical procedure of MACI began with the arthroscopic harvesting of cartilage tissue for chondrocyte cultivation. Cartilage tissue (50–150mg) was obtained from the non-weightbearing supracondylar region about the femoral condyles and placed into serum free nutrient media [17]. The isolation and cultivation of autologous chondrocytes was conducted in Australian Therapeutic Goods Administration (TGA), licensed facilities at Verigen Australia as according to the method previously described

[17]. After acceptable cell density was ascertained in vitro, cells were seeded onto the collagen membrane and transported to theatre. The surgical technique used was similar to that previously described [10]. The defect site was accessed via a parapatellar incision and arthrotomy in a tourniquet controlled field. During implantation, defects were thoroughly curetted to remove reactive fibrous tissue build-up and define defect borders. The chondrocyte seeded collagen scaffold was then shaped to match defect geometry. Bleeding was controlled by adrenalin swab. Once the scaffolds were correctly shaped, the defect was filled to the surface by the injection of Tisseel® fibrin sealant (Baxter AG, Vienna, Austria) and the shaped scaffold was press fit into the defect. Full range movement of the joint was made (4–5 times) prior to closure to assure implant stability. Any instability or delamination was countered by the reapplication of fibrin sealant, or re-shaping of the graft with excess membrane. Structured exercise sessions commenced prior to surgery in order to prepare patients physically and mentally for the rigors of surgery and the lengthy post-operative recovery. Continuous passive motion was commenced 1 day postoperatively, and the patients were gradually returned to weightbearing activity over the ensuing months by participation in a graduated rehabilitation program (Figure 2) designed for ACI recovery to protect and stimulate the healing process [23]. Following surgery, patients underwent an intensive, individually tailored MACI rehabilitation program. The underling principle for this

Page 117 program was to encourage and maximize the chondrocyte maturation process, whilst minimizing the risk of graft failure through overload or delamination.

Clinical Assessment

Six-Minute Walk Distance Test: Functional capacity and general gait function were determined by the six-minute walk test (6MWT) [24,25]. Subjects were instructed to walk as fast as possible in six minutes on a flat indoor 25m course, trying to cover the maximum distance without over-exertion. The final score was calculated as the total distance walked to the nearest 1.0m.

The Knee Injury and Osteoarthritis Outcome Score (KOOS): Subjective knee function was assessed pre- and postoperatively using the knee injury and osteoarthritis outcome score (KOOS), a knee-specific outcome instrument developed by Roos et al [26,27].

The KOOS evaluates both short-term and long-term consequences of knee injury, and is self-administered. The questionnaire comprises 42 items within five domains: pain (9 items), symptoms (7 items), function in activities of daily living (ADL; 17 items), function in sport and recreation (Sport/Rec, 5 items), and knee-related quality of life

(KQOL, 4 items).

MRI Assessment

MRI scans were conducted at 3, 12 and 24 months postoperatively using a 1.5 Tesla closed unit with extremity coil (Siemens Vision; Siemens, Erlangen, Germany), employing a previously described cartilage imaging sequence protocol and blinded evaluation by a consultant musculoskeletal radiologist [14]. Each MRI parameter

(defect infill, signal intensity, surface contour, structure, border integration, subchondral lamina, subchondral bone and effusion) was scored against a series of sample images, ranked from 1=“Poor” to 4=“Excellent” then multiplied by a weighting factor to obtain

Page 118 the final MRI composite score. MRI data was also assessed in disaggregated fashion by category in accordance to the recommendations of Marlovits et al [28,29]. Synovitis was recorded and graded separately. Intra-observer reliability assessment was conducted using 20 image pairs in which a significant (P<0.01, rho=0.787) correlation between samples was observed and no significant (P>0.01) difference was recorded between test and retest images.

Determination of Graft Failure

Graft failure was determined both clinically and radiographically. Clinically, graft failure was defined as the deterioration of the knee condition upon examination, with indicators that included the presence of mechanical symptoms such as locking, catching and/or associated knee joint pain. Radiographically, graft failure was defined by evidence of suboptimal defect infill and/or evidence of internal derangement (such as clefts, fissures, or basal delamination). Any that showed clinical and radiographical evidence of failure would be referred back to the surgeon for patient-specific management.

Statistical Analysis

Data were stored on Microsoft Excel spreadsheets and analyzed using SPSS (version

10.0) for Windows. An intention to treat analysis was performed using the ‘last value carried forward’ technique (5% of data cells), and changes between postoperative time points compared using repeated measures analysis of variance (ANOVA). Post-hoc analysis was performed using Tukey’s HSD. The Pearson Correlation Coefficient was used to analyse the relationship between MRI and functional scoring, as well as implant variables (defect area and implanted cell number) and functional scoring. All reported

P-values were two-tailed and values less than 0.05 were considered significant.

Page 119

RESULTS:

Clinical Outcomes

Statistical analysis of the functional outcomes indicated that patients experienced a significant (P<0.001) improvement in six-minute walk distance and all five KOOS subscales from preoperatively to 24 months postoperatively (Table 2). Though MACI patients demonstrated an increased distance covered in the six-minute walk test from before surgery to 24 months postoperatively, scores for this parameter were suppressed at 3 months due to the weightbearing constraints of the rehabilitation protocols. Post- hoc analysis demonstrated the improvement occurred predominantly in the first 12 months (P<0.05) and was maintained to the 24 month postoperative time point (Figure

3). Post-hoc KOOS analyses also revealed the observed improvement in knee pain, symptoms, and ADL occurred predominantly in the first 12 months post-MACI, then plateaued; whereas the improvement in sport and recreation function increased linearly from 3 to 24 months (Figure 4). The knee related quality of life subscale of the KOOS score improved significantly (P<0.05) from 3 to 12 months following surgery, then only marginally improved from 12 to 24 months (P>0.05).

MRI Outcomes

MACI patients demonstrated an increased MRI composite score over time that improved significantly from 3 to 24 months postoperatively (P<0.001). Post-hoc analysis demonstrated that improvement occurred predominantly in the first 12 months

(Figures 5 and 6), then plateaued at 24 months postoperatively.

At three months following MACI (Figure 6B), 45% (n=13) of the MACI grafts exhibited good to excellent filling of the chondral defect, with signal intensity described as good to excellent in 28% (n=8) of grafts. Good to excellent border

Page 120 integration of the reparative tissue with the adjacent cartilage was evident in 76%

(n=22) of grafts, whilst surface continuity was good to excellent in 83% (n=24) of cases. Good to excellent subchondral lamina was observed in 96% (n=28) of the cases

(indicative that it was intact at the time of surgery), and 83% (n=24) of cases exhibited good to excellent resolution of preoperative subchondral bone edema. Joint effusion was only evident in 24% (n=7) of cases, and 55% (n=16) exhibited synovitis. No graft hypertrophy was reported at 3 months.

At 12 months following MACI (Figure 6C), good to excellent filling of the defect had increased to 76% (n=22) of grafts, and signal intensity had improved to 93% (n=27).

Good to excellent border integration of reparative tissue with adjacent cartilage was seen in 79% (n=23) of cases, with intact surface continuity in 86% (n=25) of grafts.

Good to excellent restoration of the subchondral lamina was evident in all cases, and

93% (n=27) of cases showed complete resolution of subchondral bone edema. Joint effusion improved to only 3% (n=1) of cases, although 28% (n=8) of cases had persistent synovitis. Minor graft hypertrophy was reported in two cases.

By 24 months post-MACI (Figure 6D), there was no change in defect infill from the 12 month time point, and signal intensity and graft structure had achieved a good to excellent rating in 86% (n=25) of cases. Good to excellent repair integration was observed in 83% (n=24) of cases, with intact surface continuity in 83% (n=24) of grafts.

Effusion was present in only one case, and synovitis had improved to 20% (n=6). A third case exhibited graft hypertrophy at this time point.

Page 121

Outcome Correlation Analysis

Significant positive correlation between the composite MRI score and functional outcome was evidenced. MRI and 6-minute walk distance (P < 0.01, rho = 0.426), MRI and KOOS pain subscale (P < 0.01, rho = 0.301), MRI and KOOS symptoms subscale

(P < 0.01, rho = 0.422), MRI and KOOS activities of daily living subscale (P < 0.01, rho = 0.346), MRI and KOOS sport and recreation subscale (P < 0.01, rho = 0.413),

MRI and KOOS knee-related quality of life (P < 0.01, rho = 0.302) all showed a significant correlation. On the other hand, neither defect area nor implanted cell number showed a significant (P>0.05) correlation to any functional outcome parameters, suggesting that in vitro variations have little consequence on clinical outcome.

Complications

Five patients developed deep vein thrombosis (DVT), were administered anti- coagulants, and fully recovered. Four complications attributable to the MACI procedure were noted (including three cases of graft hypertrophy), however surgical intervention was not deemed necessary as all patients were asymptomatic. One patient developed severe patella tendonitis, thought to be related to a concomitant tibial tubercle transfer procedure, and was successfully managed by physiotherapy and corticosteroids. One traumatic graft delamination (partial) was detected at three months by MRI. Upon clinical review, the patient admitted severe non-compliance to rehabilitation, resulting in trauma to the graft site in the tenth postoperative week. The detached portion of the graft was arthroscopically removed and good to excellent infill was noted at 12 month

MRI.

Page 122

DISCUSSION

Matrix-induced autologous chondrocyte implantation (MACI), the third generation of

ACI, bypasses the many problems associated with periosteum as a defect cap in conventional ACI, and the need to suture the collagen membrane to the defect in collagen-covered ACI. Using the MACI technique, the periosteal patch is substituted with an inert acellular type I/III collagen membrane, hence periosteal harvesting is obviated, the operation is simplified, and anaesthetic time and incision length are reduced. Herein, we have evidenced structural and functional improvement at 2 years in patients treated for chondral knee injury with MACI; a finding consistent with previous studies assessing this treatment [10,20,21,30]. Moreover, the incidence of periosteal

ACI graft hypertrophy (requiring debridement) is generally reported in 25-36% of cases, yet only three cases (10%) of minor hypertrophic growth (no debridement required) were noted in this study [6,31]. We also noted a significant correlation between MRI score and functional outcome in the present study.

Subjective knee function and six-minute walk assessment among our MACI patients improved over time in parallel with maturation of the graft. Patients in our study experienced a significant reduction in knee pain, improvements in sports and recreation function, activities of daily living, and knee-related quality of life from 3 to 24 months postoperatively, with the majority of improvement in the first 12 months. Our 24-month

KOOS results were also comparable to those reported by Marlovits et al, thereby indicating that improvements following MACI are, for at least 2 years, durable with return to activity [28]. Noteworthy, although the capacity to walk for distance increased at 24 months, this parameter was decreased at the 3 month postoperative time point.

This was most likely related to the trauma of surgery and associated early postoperative restraints used to protect the graft [14,23]. These findings suggest that whilst structured

Page 123 rehabilitation following MACI may initially reduce patient function, over time it improves function compared to presurgery and appears to be stable to at least 2 years.

A review of the literature suggests an approximate 6% incidence of retarded knee flexion following MACI, requiring manipulation under aneasthesia (MUA).

Consequently, immobilization of the operative knee joint, irrespective of defect location, for 10-14 days postoperatively, has been advocated by some authors [21].

While in contrast, others have suggested that immobilization leads to decreased joint

ROM, followed by biological adaptation of the articular structures to the immobilized joint [23]. Indeed, we believe the biological longevity and clinical success of the graft is dependent on a controlled and graduated return to weightbearing and physical activity, and the associated biomechanical stimulation of the implanted chondrocytes. This has been evidenced at a cellular level by various studies showing the relationship between cartilage matrix synthesis and biomechanical stimuli [32-36]. Specifically, dynamic compression of cartilage stimulates matrix (aggrecan and collagen II) biosynthesis dependent on loading frequency and amplitude, whereas increased static compression

(immobilized joints) by mechanical or osmotic stress has been shown to decrease matrix

biosynthesis in a dose-dependent manner [32,35,36]. To this end, no clinical incidence of knee stiffness requiring MUA was observed in this study. Although we have not included a control group without early CPM, our clinical outcomes and lack of complication support early mobilization of the joint with CPM to improve matrix biosynthesis via controlled dynamic compression, in conjunction with rehabilitation protocols that incorporate the complexities of individual cases [23,37].

The rehabilitation protocol adopted in this study was well tolerated by the patients; however, the singular incidence of graft delamination in our cohort highlights the

Page 124 clinical importance of a protection phase coupled with patient compliance during the initial rehabilitation period. As commented by Willers et al, it is not only important to achieve successful cultivation and surgical implantation of the graft, but it is vital that the integrity of the graft be appropriately protected during all phases of the rehabilitation process to optimize repair tissue maturation [38]. Interestingly, it is noteworthy that the delamination observed at 3 months only dislodged the superficial layer of the graft, leaving residual reparative tissue intact in the base of the defect, which continued to fill the defect and mature with time. Whilst this may suggest that cell migration and integration into the defect is sufficient at 3 months to facilitate adequate repair, variation in defect maturation between cases must be appreciated.

Examination by arthroscopy or tissue biopsy is controversial because of the invasiveness to patients. Also, the high incidence of inadequate biopsy (55% as reported by ICRS [39] precludes meaningful interpretation in the majority of specimens. We consider it unethical to subject ACI patients to routine ‘second-look’ arthroscopy or biopsy when the MACI graft is considered to be progressing well clinically. Therefore, we have examined the potential for MRI assessment as a postoperative measure of graft outcome and durability. MRI evaluation of defect infill and tissue regeneration following MACI revealed a similar maturation pathway to that reported by previous studies of periosteal ACI and collagen-covered ACI [5,21,40,41]. We have shown that

MRI allows evaluation of cartilage repair infill, graft incorporation and tissue signal, surface congruity. Additionally, post-operative complications such as graft delamination, arthrofibrosis, fissure formation, subchondral bone oedema, and hypertrophy of the graft can be assessed with this technology. Thus we believe, in conjunction with the literature, that MRI accurately allows a non-invasive follow-up

Page 125 method for the detection of postoperative complications, and structural follow-up [42-

45].

The relationship between MRI outcome and functional outcome is unclear. Takahashi et al reported no correlation between clinical outcome and MRI scoring of the graft, whilst

Marlovits et al noted significant correlation between MRI scoring and clinical outcome in terms of defect filling, repair tissue structure, repair tissue signal intensity, and changes in the subchondral bone [28,46]. Consistent with these findings, the present study (and previous collagen-covered ACI study) found a significant correlation between MRI and functional outcome [14]. It must be noted however that MRI should only be used to complement clinical outcomes, as it is widely appreciated that repair structure does not reliably predict the patient’s functional restoration. Furthermore, the lack of standardization in outcome classification systems, and interobserver variability, adds confusion to any inter-study interpretation of this technology. In saying this, MRI is a valuable tool for following the postoperative integrity of the graft and the progression of the repair tissue, evaluating the possible need for revision procedures, and comparing differential treatment options. Additionally, cell number and defect area showed no significant correlation to functional outcome. This is consistent with our previous animal study, and importantly, supports the current cell cultivation protocols used in this technology, and the application of MACI to all defect sizes in the knee [15].

In summary, this study provides insight into the morphological and functional progression of the regenerative cartilage produced by MACI through the use of the latest MRI evaluation parameters and reliable functional outcome measures. The results supplement the existing clinical, radiological, and histological information, to give a better understanding of the postoperative progression of cartilage repair by MACI. As

Page 126 contributed to here, further investigation of the relationship between MRI and functional outcome following MACI is imperative in order to determine the degree to which native cartilage structure and function must be restored to achieve durable clinical results.

ACKNOWLEDGEMENTS

This study was funded by a research grant provided by The National Health and

Medical Research Council, administered by the council on behalf of the Australian

Government. Unless otherwise specified, the data given in this review is based on work carried out at the University of Western Australia.

Page 127

Figure 1: Paradigm of matrix-induced autologous chondrocyte implantation (MACI) cartilage regeneration. (1) Implantation of chondrocyte seeded membrane into the fibrin sealant-covered base of the debrided chondral defect (day of implantation). (2) Cell migration of chondrocytes from the cambium surface of the membrane into the fibrin sealant matrix. Host resorption of the collagen membrane has also commenced (2-5 days following implantation). (3) Matrix production by implanted autologous chondrocytes. Type II collagen, aggrecan and other matrix proteins important for healthy articular cartilage function are synthesized by the newly implanted cells (1-12 months following implantation). (4) Matrix maturation and hyaline-like/hyaline cartilage formation. Cartilage infill is complete, chondrocyte morphology and surrounding matrix appears healthy (or similar to surrounding native tissue) and graft cartilage is well integrated with the adjacent cartilage (12-24 months following implantation). Page 128

Figure 2: The graduated return to weightbearing administered to patients during functional rehabilitation following their MACI surgery. Gradual loading of the joint is conducted to stimulate maturation and adaptation of hyaline-like cartilage infill through physiologically induced chondrocyte biosynthesis.

Page 129

Figure 3: Improvement in six-minute walk distance to 24 months postoperatively (n = 28). An increased six-minute walk distance was noted from 6 months postoperatively compared to preoperatively. Scores for this parameter were suppressed at 3 months, and improvement occurred predominantly in the first 12 months onwards (P<0.05).

Page 130

Figure 4: Improvement in the five sub domains of KOOS to 24 months postoperatively (n = 28). Observed improvements in knee pain, symptoms, and ADL were noted predominantly in the first 12 postoperative months onwards. While improvements in sport and recreation function increased linearly from 3 months onwards, and knee related quality of life improved significantly from 3 months onwards (P>0.05). Improvement in the sport and recreation and knee related quality of life subscales were reduced compared to the other subscales due to the higher functional demand of these components. Total KOOS scores (0 = extreme knee problems and 100 = no knee problems), ADL = activities of daily living, Sport&Rec = sport and recreation function, KQOL = knee-related quality of life.

Page 131

Figure 5: Improvement in MRI composite score to 24 months postoperatively. MACI patients demonstrated an increased MRI composite score that improved significantly (P<0.001) from 3 to 24 months postoperatively, with predominant increase seen in the first 12 months. Specifically, at 12 months, good to excellent filling of the defect had increased to 76% of grafts (from 45%), and signal intensity had improved to 93% of grafts (from 28%). * P<0.001 compared to 3 months.

Figure 6: Sagittal proton density fast spin echo magnetic resonance image of a MACI graft in a patient treated for a full thickness chondral defect of the medial femoral condyle. Compared to preoperatively (A), the MACI graft was hyperintense and of reduced thickness compared with the adjacent normal articular cartilage three months postoperatively (B). At one year postoperatively (C) the MACI graft displayed a heterogeneous appearance with similar thickness to the adjacent normal cartilage. Reconstitution of the subchondral bone plate (arrows) improved markedly from the 3 month time point. At two years postoperatively, the MACI graft remained intact and demonstrated a heterogeneous graft signal compared to the adjacent native cartilage. Border integration was smooth, with no radiographical evidence of fissures or clefts between the native cartilage or within the graft, and restoration of the subchondral plate appeared almost complete. Page 132 Page 133

Table 1: Surgical histories for the MACI cohort. Procedures conducted before, during, and after matrix-induced autologous chondrocyte implantation (MACI).

Page 134

Table 2: Statistics summary for the six-minute walk test, and five subscales of the KOOS score (pain, symptoms, activities of daily living, sport and recreation function, and knee related quality of life).

Page 135

REFERENCES

1. Brittberg M, Lindahl A, Nilsson A, Ohlsson C, Isaksson O, Peterson L. Treatment of deep cartilage defects in the knee with autologous chondrocyte transplantation. N Engl J Med 1994;331(14):889-95. 2. Peterson L, Menche D, Grande D. Chondrocyte transplantation - an experimental rabbit in the rabbit. In: 30th Annual Orthopaedic Research Society; 1984; Atlanta: Orthopaedic Research Society; 1984. p. 218. 3. Driesang IM, Hunziker EB. Delamination rates of tissue flaps used in articular cartilage repair. J Orthop Res 2000;18(6):909-11. 4. Henderson I, Tuy B, Oakes B. Reoperation after autologous chondrocyte implantation. Indications and findings. J Bone Joint Surg Br 2004;86(2):205-11. 5. King PJ, Bryant T, Minas T. Autologous chondrocyte implantation for chondral defects of the knee: indications and technique. J Knee Surg 2002;15(3):177-84. 6. Minas T. Autologous chondrocyte implantation for focal chondral defects of the knee. Clin Orthop Relat Res 2001(391 Suppl):S349-61. 7. Ueno T, Kagawa T, Mizukawa N, Nakamura H, Sugahara T, Yamamoto T. Cellular origin of endochondral ossification from grafted periosteum. Anat Rec 2001;264(4):348-57. 8. Wood JJ, Malek MA, Frassica FJ, Polder JA, Mohan AK, Bloom ET, Braun M, Cote TR. Autologous cultured chondrocytes: adverse events reported to the United States Food and Drug Administration. J Bone Joint Surg Am 2006;88(3):503-7. 9. Briggs TW, Mahroof S, David LA, Flannelly J, Pringle J, Bayliss M. Histological evaluation of chondral defects after autologous chondrocyte implantation of the knee. J Bone Joint Surg Br 2003;85(7):1077-83. 10. Cherubino P, Grassi FA, Bulgheroni P, Ronga M. Autologous chondrocyte implantation using a bilayer collagen membrane: a preliminary report. J Orthop Surg (Hong Kong) 2003;11(1):10-5. 11. Gooding CR, Bartlett W, Bentley G, Skinner JA, Carrington R, Flanagan A. A prospective, randomised study comparing two techniques of autologous chondrocyte implantation for osteochondral defects in the knee: Periosteum covered versus type I/III collagen covered. Knee 2006;13(3):203-10. 12. Kirilak Y, Pavlos NJ, Willers CR, Han R, Feng H, Xu J, Asokananthan N, Stewart G, Henry P, Wood D, Zheng M.H. Fibrin sealant promotes migration

Page 136

and proliferation of human articular chondrocytes: Possible involvement of thrombin and protease-activated receptors. Int J Mol Med 2006;17(4):551-8. 13. Krishnan SP, Skinner JA, Carrington RW, Flanagan AM, Briggs TW, Bentley G. Collagen-covered autologous chondrocyte implantation for osteochondritis dissecans of the knee: TWO- TO SEVEN-YEAR RESULTS. J Bone Joint Surg Br 2006;88(2):203-5. 14. Robertson WB, Fick D, Wood DJ, Linklater JM, Zheng MH, Ackland TR. MRI and clinical evaluation of collagen-covered autologous chondrocyte implantation (CACI) at two years. Knee 2007;14(2):117-27. 15. Willers C, Chen J, Wood D, Xu J, Zheng MH. Autologous chondrocyte implantation with collagen bioscaffold for the treatment of osteochondral defects in rabbits. Tissue Eng 2005;11(7-8):1065-76. 16. Sohn DH, Lottman LM, Lum LY, et al. Effect of gravity on localization of chondrocytes implanted in cartilage defects. Clin Orthop Relat Res 2002(394):254-62. 17. Zheng MH, Willers C, Kirilak L, Yates P, Xu J, Wood D, Shimmin A. Matrix- induced autologous chondrocyte implantation (MACI): biological and histological assessment. Tissue Eng 2007;13(4):737-46. 18. Ronga M, Grassi FA, Bulgheroni P. Arthroscopic autologous chondrocyte implantation for the treatment of a chondral defect in the tibial plateau of the knee. Arthroscopy 2004;20(1):79-84. 19. Basad E, Stürz H, Steinmeyer J. Treatment of chondral defects with MACI or microfracture. First results of a comparative clinical study [Die behandlung chondraler defekte mit MACI oder microfracture - erste Ergebnisse einer vergleichenden klinischen Studie]. Orthopädische Praxis 2004;40:6-10. 20. Behrens P, Bitter T, Kurz B, Russlies M. Matrix-associated autologous chondrocyte transplantation/implantation (MACT/MACI)--5-year follow-up. Knee 2006;13(3):194-202. 21. Bartlett W, Skinner JA, Gooding CR, et al. Autologous chondrocyte implantation versus matrix-induced autologous chondrocyte implantation for osteochondral defects of the knee: a prospective, randomised study. J Bone Joint Surg Br 2005;87(5):640-5. 22. Outerbridge RE. The etiology of chondromalacia patellae. J Bone Joint Surg Br 1961;43-B:752-7.

Page 137

23. Hambly K, Bobic V, Wondrasch B, Van Assche D, Marlovits S. Autologous Chondrocyte Implantation Postoperative Care and Rehabilitation: Science and Practice. Am J Sports Med 2006. 24. laboratories. Acopsfcpf. ATS statement: guidelines for the six minute walk test. Am J Respir Crit Care Med 2002;166(1):111-7. 25. Enright PL. The six-minute walk test. Respir Care 2003;48(8):783-5. 26. Roos EM, Roos HP, Lohmander LS, Ekdahl C, Beynnon BD. Knee Injury and Osteoarthritis Outcome Score (KOOS)--development of a self-administered outcome measure. J Orthop Sports Phys Ther 1998;28(2):88-96. 27. Roos EM, Lohmander LS. The Knee injury and Osteoarthritis Outcome Score (KOOS): from joint injury to osteoarthritis. Health Qual Life Outcomes 2003;1(1):64. 28. Marlovits S, Singer P, Zeller P, Mandl I, Haller J, Trattnig S. Magnetic resonance observation of cartilage repair tissue (MOCART) for the evaluation of autologous chondrocyte transplantation: determination of interobserver variability and correlation to clinical outcome after 2 years. Eur J Radiol 2006;57(1):16-23. 29. Marlovits S, Striessnig G, Resinger CT, et al. Definition of pertinent parameters for the evaluation of articular cartilage repair tissue with high-resolution magnetic resonance imaging. Eur J Radiol 2004;52(3):310-9. 30. Trattnig S, Ba-Ssalamah A, Pinker K, Plank C, Vecsei V, Marlovits S. Matrix- based autologous chondrocyte implantation for cartilage repair: noninvasive monitoring by high-resolution magnetic resonance imaging. Magn Reson Imaging 2005;23(7):779-87. 31. Gooding CR, Bartlett W, Bentley G, Skinner JA, Carrington R, Flanagan A. A prospective, ranomised study comparing two techniques of autologous chondrocyte implantation for osteochondral defects in the knee: Periosteum covered versus type I/III collagen covered. Knee 2006. 32. Burton-Wurster N, Vernier-Singer M, Farquhar T, Lust G. Effect of compressive loading and unloading on the synthesis of total protein, proteoglycan, and fibronectin by canine cartilage explants. J Orthop Res 1993;11(5):717-29. 33. Burton-Wurster N, Mateescu RG, Todhunter RJ, et al. Genes in canine articular cartilage that respond to mechanical injury: gene expression studies with Affymetrix canine GeneChip. J Hered 2005;96(7):821-8.

Page 138

34. Fitzgerald JB, Jin M, Dean D, Wood DJ, Zheng MH, Grodzinsky AJ. Mechanical compression of cartilage explants induces multiple time-dependent gene expression patterns and involves intracellular calcium and cyclic AMP. J Biol Chem 2004;279(19):19502-11. 35. Sah RL, Kim YL, Grodzinsky AJ, Plaas AHK, Sandy JD. Effects of static and dynamic compression on cartilage metabolism in cartilage explants. New York: Raven Press; 1992. 36. Torzilli PA, Grigiene R, Huang C, et al. Characterization of cartilage metabolic response to static and dynamic stress using a mechanical explant test system. J Biomech 1997;30(1):1-9. 37. Robertson WB GH, Ackland T. Standard Practice Exercise Rehabilitation Protocols for Matrix Induced Autologous Chondrocyte Implantation Femoral Condyles. Hollywood Functional Rehabilitation Clinic Publications 2004. 38. Willers C, Partsalis T, Zheng MH. Articular cartilage repair: procedures versus products. Expert Rev Med Devices 2007;4(3):373-92. 39. Mainil-Varlet P, Aigner T, Brittberg M, et al. Histological assessment of cartilage repair: a report by the Histology Endpoint Committee of the International Cartilage Repair Society (ICRS). J Bone Joint Surg Am 2003;85- A Suppl 2:45-57. 40. Henderson IJ, Tuy B, Connell D, Oakes B, Hettwer WH. Prospective clinical study of autologous chondrocyte implantation and correlation with MRI at three and 12 months. J Bone Joint Surg Br 2003;85(7):1060-6. 41. James SL, Connell DA, Saifuddin A, Skinner JA, Briggs TW. MR imaging of autologous chondrocyte implantation of the knee. Eur Radiol 2006:1-9. 42. Bachmann G, Basad E, Lommel D, Steinmeyer J. [MRI in the follow-up of matrix-supported autologous chondrocyte transplantation (MACI) and microfracture]. Radiologe 2004;44(8):773-82. 43. Chung CB, Frank LR, Resnick D. Cartilage imaging techniques: current clinical applications and state of the art imaging. Clin Orthop Relat Res 2001(391 Suppl):S370-8. 44. Polster J, Recht M. Postoperative MR evaluation of chondral repair in the knee. Eur J Radiol 2005;54(2):206-13. 45. Roberts S, McCall IW, Darby AJ, et al. Autologous chondrocyte implantation for cartilage repair: monitoring its success by magnetic resonance imaging and histology. Arthritis Res Ther 2003;5(1):R60-73.

Page 139

46. Takahashi T, Tins B, McCall IW, Richardson JB, Takagi K, Ashton K. MR appearance of autologous chondrocyte implantation in the knee: correlation with the knee features and clinical outcome. Skeletal Radiol 2006;35(1):16-26.

Chapter 8

Clinical MACI: Patient Satisfaction Survey

Thesis publication #7: Willers C, Zheng MH. Matrix-induced Autologous Chondrocyte Implantation (MACI®): A Retrospective Survey of 202 Cases.

Page 140

STATEMENTS OF AUTHORSHIP CONTRIBUTION FROM CO-AUTHORS

Matrix-induced Autologous Chondrocyte Implantation (MACI®): A Retrospective Survey of 202 Cases

Willers C, Zheng MH

Willers C (PhD Candidate) Major contribution to the planning, execution, analysis, and interpretation of all research. Major contribution to writing of the manuscript.

Signature of Principal Author: ……………………………… Date: .………………

Zheng MH (Supervisor) Major contribution to the planning and interpretation of the research. Minor contribution to writing of the manuscript.

Signature of Co-Author: ……………………..……………… Date: …………………

Page 141

STATEMENTS OF PERMISSION FROM CO-AUTHORS

I, Ming-Hao Zheng give permission to Craig Willers, principal author of the paper: Matrix-induced Autologous Chondrocyte Implantation (MACI®): A Retrospective Survey of 202 Cases, to include this paper as Chapter 8 of his PhD thesis entitled Matrix-induced Autologous Chondrocyte Implantation for Articular Cartilage Injury: Biology, Histology, and Clinical Outcomes.

Signature: …………………………………. Date: …………………………………. Page 142

Matrix-induced Autologous Chondrocyte Implantation (MACI®): A Retrospective Survey of 202 Cases.

Craig Willers BSc(H1) M(Med)Sc, Ming-Hao Zheng PhD DM FRCPath

Centre for Orthopaedic Research, School of Pathology and Surgery, University of Western Australia, 2nd Floor M-block QEII Medical Centre, Nedlands, Perth WA, 6009, Australia

Mr Craig Willers: TELEPHONE: 08 93463213, FAX: 08 93463210, EMAIL: [email protected] Prof. Ming-Hao Zheng: TELEPHONE: 08 93463213, FAX: 08 93463210, EMAIL: [email protected]

Corresponding Author: Professor Ming-Hao Zheng Director of Research Department of Orthopaedics, School of Pathology and Surgery University of Western Australia 2nd Floor M-block QEII Medical Centre Nedlands 6009, Australia Telephone: 08 9346 3213 Facsimile: 08 9346 3210 E-mail: [email protected]

Page 143

ABSTRACT

Matrix-induced autologous chondrocyte implantation (MACI®) is becoming increasingly popular for the treatment of large articular cartilage defects, however studies assessing this technique generally involve small cohorts complicated by a lack of outcome score standardization. We have conducted a retrospective multi-centre cohort study of 202 patients using a single ten-question questionnaire covering patient symptoms, function, quality of life, and satisfaction (based largely on the Lysholm and

Cincinnati scales). The mean follow-up was 22.7±8.6 months, mean age 36.9±10.7 years, and mean size of defect was 4.8cm2. Fifty-nine percent of patients had previous surgery to their knee. Overall results showed that 167 (83%) of the surveyed patients had a good to excellent MACI outcome. We have also found significant improvement in younger patients (<30 vs. 30-50, P=0.001), those surveyed after more than 24 months

(>24 months vs. 12-24 months, P=0.03), those participating in formalised rehabilitation

(with vs. without, P=0.007), and those with defects located on the femoral components of the knee joint (trochlea vs. patella, P=0.007; medial femoral condyle vs, patella,

P=0.03). No significant difference (P>0.05) in outcome was observed with variations in cell cultivation parameters. MACI® showed a high rate of good/excellent patient satisfaction and clinical outcome, with better results in younger, more active knee joints.

This study suggests that MACI® produces good to excellent outcomes in satisfied patients, but may be improved by postoperative conditioning of the joint.

Page 144

INTRODUCTION

Damaged articular cartilage is a common complaint observed in orthopaedic clinics. A large cohort study from Curl et al (1997) in the United States has shown 63% of 31,516 knee arthroscopies revealed chondral lesions and 20% had full-thickness lesions, with

2.7 lesions per knee on average, mainly in the medial femoral condyle [1].

Fibrocartilage is produced by native repair mechanisms after injury to the articular surface, but this tissue has biomechanical characteristics inferior to that of native hyaline articular cartilage, and inevitably translates to short-term symptomatic relief only [2]. Initially, arthroscopic lavage, abrasion arthroplasty, subchondral drilling, microfracture, and osteochondral autografting were trialed as surgical cartilage repair paradigms, but produced variable outcomes and were prone to complication [3-8]. The shortcomings of these conventional treatments have spawned the development of cell- based cartilage repair technologies such as autologous chondrocyte implantation (ACI).

Peterson et al presented the first evidence in 1984 that ACI was a credible treatment for chondral injury [9], documenting successful transplantation of autologous chondrocytes under a periosteal flap into cartilage defects created in rabbit patellae. This allowed the approval of human trials, which culminated in the 1994 publication by Brittberg et al on the success of periosteal ACI (ACI-P) for treating cartilage defects in 23 patients [10].

Brittberg reported 87% good or excellent results in femoral condylar repair and 73% demonstration of hyaline-like cartilage upon microscopy. Moreover, although Knutsen et al initially reported improved clinical outcomes but no histological difference in microfracture patients compared to ACI, more recent research has found comparable clinical outcome and significant histological improvement in ACI patients compared to microfracture [11, 12]. However, Minas has reported that complications linked with the

Page 145 use of periosteum in ACI-P may occur in 20% to 25% of cases; with others reporting up to 87% lifting or delamination of the periosteal graft [13, 14].

A new generation ACI, matrix-induced autologous chondrocyte implantation (MACI®), addresses the aforementioned concerns of ACI-P by substituting the periosteum with an inert type I/III collagen membrane. As the cells are directly seeded onto the membrane then adhered to the defect base using fibrin sealant, there is no periosteum requirement, and hence risk of injected cell solution leakage, or micro-defects created by suturing the graft to the defect. The MACI® technique was first clinically illustrated in 2003 [15].

This report published 6-month postoperative results stating no complications, improved clinical and functional outcomes, and MRI-visualized hyaline-like cartilage.

While it is widely suggested that postoperative care and rehabilitation may improve the clinical outcomes of ACI, no investigations have proven this theory. Compounding this issue, Jakobsen et al recently reported on the quality of 61 cartilage repair studies, finding several methodological deficiencies [16]. In particular, a total of 27 clinical outcome measurements were used to assess outcome, of which none had been validated for use in patients with cartilage injuries. Because of this lack of measurement standardization between groups, large-scale multi-centre cohort analysis is difficult to accumulate. Hence, albeit retrospective, the objective of this study was to evaluate the function, quality of life and satisfaction of a large multi-centre MACI® cohort using a single patient questionnaire, and to assess the effect of pre- and post-operative variables on clinical outcome.

Page 146

METHODS

The surgical procedure of MACI began with the arthroscopic harvesting of cartilage tissue for chondrocyte cultivation. Cartilage tissue (50–150mg) was obtained from the non-weightbearing supracondylar region about the femoral condyles and placed into serum free nutrient media. The isolation and cultivation of autologous chondrocytes was conducted in Australian Therapeutic Goods Administration (TGA) licensed facilities at

Verigen Australia as according to the method previously described. After acceptable cell density was ascertained in vitro, cells were seeded onto the collagen membrane

(minimum cell density of 2.5x105cells/cm2) and transported to theatre. The surgical technique used was similar to that previously described [15]. The defect site was accessed via a parapatellar incision and arthrotomy in a tourniquet controlled field.

During implantation, defects were thoroughly curetted to remove reactive fibrous tissue build-up and define defect borders. The chondrocyte seeded collagen scaffold was then shaped to match defect geometry. Bleeding was controlled by adrenalin swab. Once the scaffolds were correctly shaped, the defect was filled to the surface by the injection of

Tisseel® fibrin sealant (Baxter AG, Vienna, Austria) and the shaped scaffold was press fit into the defect. Full range movement of the joint was made (4–5 times) prior to closure to assure implant stability. Any instability or delamination was countered by the reapplication of fibrin sealant, or re-shaping of the graft.

Two hundred and forty-two patients were recruited by 11 surgeons who performed the

MACI® procedure between early 2002 and early 2006. The project was approved by the institution’s Human Research Ethics Committee. Patient consent was obtained through mail or verbal communication via the surgeon. No surgeon or patient received any type of financial incentive to participate in this survey. The survey was conducted independently without influence from the surgeon or company. Upon the acquisition of

Page 147 consent, patient contact details and surgical histories were given to the lead author who was the sole handler of all subsequent patient information. The original patient group was narrowed to 202 patients (16% reduction) for the survey. Patients were included if they were 12 months or greater postoperatively following MACI® to the knee joint.

Patient exclusion was based on significant reoperation of the joint following MACI®

(unrelated realignment osteotomy etc), significant cognitive impairment, or if they were not contactable.

The questionnaire was comprised of ten questions, with each explained to the patient in detail to assure comprehension. As Table 1 shows, the questions were based mainly on pain and symptom relief, functional restoration of the joint, quality of life, and patient satisfaction with MACI®. In addition to the raw answers to the questionnaire, the answers were converted to a point scale to gain an overall satisfaction score, with point weighting based on those used in the Lysholm and Cincinnati outcome measures [17,

18]. Scores ranged from 0 to 92. Overall scores were then rated as excellent (>80), good

(55-79), fair (30-54), or poor (<30), based on the Cincinnati rating system.

Rehabilitation status was not included in the overall satisfaction score. Once overall satisfaction was calculated, various sub-groups were demarcated for the analysis of variables within the surveyed cohort. Statistical significance between clinical variables was examined within postoperative months at survey (12-24 months vs. over 24 months), rehabilitation participation (with rehab vs. without rehab), patient age (under

30 years-old vs. 30-50 years-old vs. over 50 years-old), defect location (medial femoral condyle vs. lateral femoral condyle vs. trochlea vs. patella), patient gender, number of previous procedures (≥1/ ≥3 previous procedure vs. no previous procedures), procedures at implant (none vs. ≥1), procedures after implant (none vs. ≥1), defect area (≤2.25cm2 vs. ≥5cm2; only single defect cases), defect number (single defect cases vs. multiple

Page 148 defect cases), and opposing defect status (non-opposing defect cases vs. opposing defect cases). While to study the impact of the chondrocyte processing on clinical outcome, the weight of biopsy used for chondrocyte isolation (<100mg vs. >200mg), total cell

6 number inoculated onto the collagen membrane for implantation (<10 x 10 vs. >20 x

106 cells), and the total number of chondrocyte passage (≤3 passages vs. ≥5 passages) were compared to the clinical score.

All questionnaire data was statistically analysed using Microsoft Excel (Seattle, USA) and version 12.0 SPSS software (Chicago, USA). All raw questionnaire was presented as a percentage of the total patient cohort (n=202). Statistical differences between defined patient subgroups were compared using Analysis of Variance (ANOVA) and

Student t-tests. Significance was determined as P<0.05.

RESULTS

The mean age of the surveyed cohort was 36.9±10.7 years (range 14-77), the male:female ratio was 114:88, and mean time from MACI® to survey was 22.7±8.6 months (range 12-49). The mean size of isolated defects was 4.8cm2 (range 0.5-25), whilst the overall (multiple defects added cumulatively) mean size was 6.6cm2. In terms of joint biomechanics, most patients had normal tibiofemoral alignment (88%) and patellofemoral tracking (89%). Further details of patient procedures before, during, and after MACI® are detailed in Table 2. In short, the majority of previous procedures were debridement, removal of loose bodies (ROLB), and meniscectomy; procedures at implant were mainly high tibial osteotomy (HTO), tibial tubercle transfer (TTT) and anterior cruciate ligament (ACL) reconstruction; and procedures following MACI®

(only in 10% of all cases) were mainly debridement and the removal of implanted metal. Notably, 59% of patients had unsuccessful procedures prior to MACI, whereas

Page 149 only 10% of patients required surgery following MACI (most of which were unrelated to the MACI graft).

Importantly, of the overall 202 patients treated by MACI®, 167 (83%) had an excellent or good outcome. The full questionnaire outcomes can be seen in Table 3, however a summary of these outcomes follows. In terms of overall pain relief, 85% of patients rated their pain relief following MACI® as good/excellent. For their ability to perform daily activities, 81% of patients rated their improvement following MACI® as good/excellent. For return to sporting activity, 94% of patients stated they had the ability to participate in sport following their MACI, with 84% of these patients stating that MACI improved their sporting ability and 83% rating this improvement as good/excellent. In terms of general walking pain, 66% of surveyed patients reported no pain after MACI® when symptomatic before, whilst 26% of patients reported noticeable pain both before and after MACI®. In summarising the impact of MACI® on stair- climbing capacity, 73% of patients reported an improvement, 83% reported improvement from average to good/excellent, and 73% reported improvement from poor to good/excellent. In regards to swelling, 64% of patients reported no swelling after MACI®, whilst in those who reported swelling, 74% stated they were only symptomatic after strenuous exercise. Furthermore, 81% of patents reported no occurrence of locking/catching of the MACI® joint. When asked if they would undergo the MACI® procedure again if injured, 86% of patients answered ‘yes’. Interestingly, of those that stated they wouldn’t go through MACI® again if injured, over 90% stated the socioeconomic burden of rehabilitation as their primary reason. In terms of the overall satisfaction of their MACI® outcome, 82% answered good/excellent. And finally, when asked if they had participated in a formalised rehabilitation program following their

MACI® procedure, 75% answered ‘yes’.

Page 150

As detailed in Table 4, various sub-grouped variables were defined from the overall 202 patient cohort and statistically compared to investigate possible relationships between these variables and overall satisfaction score. Statistical significance was found between: postoperative months (12-24 months vs. over 24 months, P=0.03), rehabilitation participation (with rehab vs. without rehab, P=0.007), patient age (under

30 years-old vs. 30-50 years-old, P=0.001), and defect location (trochlea vs. patella,

P=0.007; medial femoral condyle (MFC) vs, patella, P=0.03). Variables not showing any significant difference(s) included patient gender, number of previous procedures, procedures at implant, procedures after implant, defect area, defect number, and opposing defect status.

In terms of chondrocyte processing, no significant difference was observed when clinical score was compared with biopsy weight (P=0.28), implanted chondrocyte density (P=0.31), or number of passage at implant (P=0.13).

DISCUSSION

This study has used a retrospective patient questionnaire to assess a large multi-centre cohort of patients having undergone MACI® for knee cartilage injury. With the standardization of using the same questionnaire for all patients, we have been able to directly compare subgroups within our cohort. Data from the present study suggests that MACI® provides good to excellent pain relief, restoration of quality of life, and function in the majority of patients. The current study has also evidenced significant differences in MACI® outcomes with age, postoperative months at survey, rehabilitation participation, and defect location within the joint. Such comparative analysis is essential to further our understanding of the factors that influence patient outcomes in the treatment of cartilage injury by cell-based repair techniques such as MACI®.

Page 151

Studies published to date have assessed the clinical outcomes of MACI® under various conditions. To this end, a prospective randomised study by Bartlett et al showed that

72% of MACI® patients exhibited good or excellent clinical results after 1 year compared to only 59% in ACI-C patients [19]. Mid-term, MACI® has recently been evidenced to produce better or much better knee function in 8 of 11 patients at 5 years follow-up, although the outcome measurements used in this study were somewhat ambiguous [20]. Similar to these findings, we have reported 83% good or excellent results (function and satisfaction) within our multi-centre cohort using a standardized questionnaire; further supporting the effectiveness of MACI® in knee cartilage repair.

We have also reported a 5% incidence of reoperation related to the cartilage surface.

This is considerably less than the 20% to 25% complication rate associated with periosteal ACI [14].

Besides the raw survey findings, a number of cohort variables were found to significantly influence the outcome of the patients. In particular, postoperative months at survey, participation in rehabilitation, patient age, and defect location illustrated a significant effect on patient outcomes. Firstly, the improvement in outcome seen in patients surveyed at more than 24 months follow-up (compared to 12-24 months).

Whilst the qualitative infill of cartilage repair tissue within articular lesions is time- variable [21], many patients may not see complete functional maturation of their repair cartilage until normal weightbearing to the knee is returned (generally 9-12 months).

Patient outcomes were also significantly improved in patients less than 30 years of age,

MFC and trochlea defects, and patient participation in a formalised rehabilitation program following their MACI. These phenomenons may be attributable to improved cartilage repair quality through increased biomechanical stimulation of the implanted

Page 152 chondrocytes following the progressive filling of the defect by the implanted chondrocytes. However, it should be noted that wound recovery and restoration of quadriceps function are also important factors in this process. To this end, younger patients generally recover faster and are involved in more physical activity and sport;

MFC and trochlea defects are subjected to greater shear and dynamic compression forces, and formalised rehabilitation offers a controlled return to full weightbearing and restoration of quadriceps function, whilst protecting the integrity of the graft. At a cellular level, this biomechanical conditioning of the cartilage translates into greater stimulation of chondrocyte matrix biosynthesis [22-24]. The importance of biomechanical conditioning in cartilage repair has also been recently clinically illustrated by Kreuz et al, who reported significantly improved knee function (ICRS and Cincinnati scores) from 6 months postoperatively in patients involved in regular or competitive sports preoperatively compared to those with rare or no sports involvement

[25]. Together with the current study, the literature suggests a clinical benefit to postoperative rehabilitation and return to sport following autologous chondrocyte therapies such as MACI®.

Interestingly, we found no correlation between chondrocyte processing variables and overall clinical score. The lack of difference in clinical outcome with implanted cell density observed is consistent with the finding of our previous animal study that realistic variation in density has no effect on defect repair histology [26]. Therefore, implantations conducted at minimum cell density (2.5x105cells/cm2) under manufacturing standards, have no differential effect on clinical outcome compared to larger seeding densities. However, the insignificant difference in score with increasing passage number is interesting given the number of publications suggesting that chondrocytes cease hyaline-specific matrix synthesis with cultivation [27-29]. Both

Page 153 these findings suggest that cellular variations seen during in vitro cultivation may not translate into the clinical setting.

The strengths of this study are its large sample size and the single outcome measure used. As noted by Jakobsen et al in quality analysis of 61 cartilage repair studies, a total of 27 clinical outcome scales were used in these studies to assess outcome [16]. Our use of a single outcome measure allowed more reliable statistical assessment of the raw outcomes and covariant analysis. Whilst our questionnaire has not been validated for use on patients with cartilage injury, as noted by Jakobsen, neither have the other major outcome measurements commonly published [16]. The most notable limitations of this study were the lack of preoperative data and the short- to mid-term follow up period.

Although patients were followed as far as 4 years in our study, a 10 to 15 year follow up is required to make definitive statements regarding the long-term efficacy of MACI®.

Also, whilst the minimum 12 month time point may be considered short, many patients are active by 9-12 months hence the rationale for their inclusion. Possibly the main limitation of this study however, was the lack of preoperative data for the cohort. Whilst most of the questions in the survey were worded to gain an idea of improvement as a result of MACI®, the questionnaire was inevitably retrospective and a prospective randomised controlled study design with preoperative data and progressive postoperative time-point analysis would have been preferable.

In summary, we have used a standardised patient questionnaire to assess improvements in symptoms, function and quality of life following MACI® and identify significant variables within a large multi-centre cohort. The present study has shown (in 202 cases) that MACI® provides good to excellent pain relief, and restoration of quality of life and function in 83% of patients. We have also evidenced significant differences in MACI®

Page 154 outcomes in younger patients, patients surveyed after more than 24 postoperative months, patients participating in formalised rehabilitation, and patients with defects located on the femoral components of the knee joint. Currently, we believe matrix- induced autologous chondrocyte implantation (MACI®) to be, at least in the mid-term, a safe and efficacious treatment option for knee cartilage injury given our current findings.

Table 1: Ten Components of the Patient Satisfaction Questionnaire.

Page 155 Page 156

Table 2: Patient Procedures Before, During, and After MACI®.

Page 157

Table 3: Raw Questionnaire Outcomes.

Page 158

Table 4: Statistical Analysis of Demographic, Pre- and Post-operative Variables with Overall Satisfaction Scores.

Page 159

REFERENCES

1. Curl WW, Krome J, Gordon ES, Rushing J, Smith BP, Poehling GG. Cartilage injuries: a review of 31,516 knee arthroscopies. Arthroscopy 1997;13(4):456-60. 2. Akeson WH, Bugbee W, Chu C, Giurea A. Differences in mesenchymal tissue repair. Clin Orthop Relat Res 2001(391 Suppl):S124-41. 3. Bentley G, Biant LC, Carrington RW, et al. A prospective, randomised comparison of autologous chondrocyte implantation versus mosaicplasty for osteochondral defects in the knee. J Bone Joint Surg Br 2003;85(2):223-30. 4. Hangody L, Fules P. Autologous osteochondral mosaicplasty for the treatment of full-thickness defects of weight-bearing joints: ten years of experimental and clinical experience. J Bone Joint Surg Am 2003;85-A Suppl 2:25-32. 5. Jackson R. Arthroscopic treatment of degenrative arthritis. New York: Raven Press; 1991. 6. Kreuz PC, Steinwachs MR, Erggelet C, et al. Results after microfracture of full- thickness chondral defects in different compartments in the knee. Osteoarthritis Cartilage 2006;14(11):1119-25. 7. Louisia S, Beaufils P, Katabi M, Robert H. Transchondral drilling for osteochondritis dissecans of the medial condyle of the knee. Knee Surg Sports Traumatol Arthrosc 2003;11(1):33-9. 8. Steadman JR, Briggs KK, Rodrigo JJ, Kocher MS, Gill TJ, Rodkey WG. Outcomes of microfracture for traumatic chondral defects of the knee: average 11-year follow-up. Arthroscopy 2003;19(5):477-84. 9. Peterson L, Menche D, Grande D. Chondrocyte transplantation - an experimental rabbit in the rabbit. In: 30th Annual Orthopaedic Research Society; 1984; Atlanta: Orthopaedic Research Society; 1984. p. 218. 10. Brittberg M, Lindahl A, Nilsson A, Ohlsson C, Isaksson O, Peterson L. Treatment of deep cartilage defects in the knee with autologous chondrocyte transplantation. N Engl J Med 1994;331(14):889-95. 11. Saris DB, Vanlauwe J, Victor J, et al. Characterized chondrocyte implantation results in better structural repair when treating symptomatic cartilage defects of the knee in a randomized controlled trial versus microfracture. Am J Sports Med 2008;36(2):235-46.

Page 160

12. Knutsen G, Engebretsen L, Ludvigsen TC, et al. Autologous chondrocyte implantation compared with microfracture in the knee. A randomized trial. J Bone Joint Surg Am 2004;86-A(3):455-64. 13. Henderson I, Tuy B, Oakes B. Reoperation after autologous chondrocyte implantation. Indications and findings. J Bone Joint Surg Br 2004;86(2):205-11. 14. Minas T. Autologous chondrocyte implantation for focal chondral defects of the knee. Clin Orthop Relat Res 2001(391 Suppl):S349-61. 15. Cherubino P, Grassi FA, Bulgheroni P, Ronga M. Autologous chondrocyte implantation using a bilayer collagen membrane: a preliminary report. J Orthop Surg (Hong Kong) 2003;11(1):10-5. 16. Jakobsen RB, Engebretsen L, Slauterbeck JR. An analysis of the quality of cartilage repair studies. J Bone Joint Surg Am 2005;87(10):2232-9. 17. Noyes FR, Mooar PA, Matthews DS, Butler DL. The symptomatic anterior cruciate-deficient knee. Part I: the long-term functional disability in athletically active individuals. J Bone Joint Surg Am 1983;65(2):154-62. 18. Tegner Y, Lysholm J. Rating systems in the evaluation of knee ligament injuries. Clin Orthop Relat Res 1985(198):43-9. 19. Bartlett W, Skinner JA, Gooding CR, et al. Autologous chondrocyte implantation versus matrix-induced autologous chondrocyte implantation for osteochondral defects of the knee: a prospective, randomised study. J Bone Joint Surg Br 2005;87(5):640-5. 20. Behrens P, Bitter T, Kurz B, Russlies M. Matrix-associated autologous chondrocyte transplantation/implantation (MACT/MACI)--5-year follow-up. Knee 2006;13(3):194-202. 21. Trattnig S, Ba-Ssalamah A, Pinker K, Plank C, Vecsei V, Marlovits S. Matrix- based autologous chondrocyte implantation for cartilage repair: noninvasive monitoring by high-resolution magnetic resonance imaging. Magn Reson Imaging 2005;23(7):779-87. 22. Marlovits S, Tichy B, Truppe M, Gruber D, Schlegel W. Collagen expression in tissue engineered cartilage of aged human articular chondrocytes in a rotating bioreactor. Int J Artif Organs 2003;26(4):319-30. 23. Mauck RL, Wang CC, Oswald ES, Ateshian GA, Hung CT. The role of cell seeding density and nutrient supply for articular cartilage tissue engineering with deformational loading. Osteoarthritis Cartilage 2003;11(12):879-90.

Page 161

24. Waldman SD, Spiteri CG, Grynpas MD, Pilliar RM, Hong J, Kandel RA. Effect of biomechanical conditioning on cartilaginous tissue formation in vitro. J Bone Joint Surg Am 2003;85-A Suppl 2:101-5. 25. Kreuz PC, Steinwachs M, Erggelet C, et al. Importance of Sports in Cartilage Regeneration After Autologous Chondrocyte Implantation: A Prospective Study With a 3-Year Follow-up. Am J Sports Med 2007. 26. Willers C, Chen J, Wood D, Xu J, Zheng MH. Autologous chondrocyte implantation with collagen bioscaffold for the treatment of osteochondral defects in rabbits. Tissue Eng 2005;11(7-8):1065-76. 27. Benya PD, Shaffer JD. Dedifferentiated chondrocytes reexpress the differentiated collagen phenotype when cultured in agarose gels. Cell 1982;30(1):215-24. 28. Zheng MH, King E, Kirilak Y, et al. Molecular characterisation of chondrocytes in autologous chondrocyte implantation. Int J Mol Med 2004;13(5):623-8. 29. Lee J, Lee E, Kim HY, Son Y. Comparison of articular cartilage with costal cartilage in initial cell yields, degree of dedifferentiation during expansion, and their redifferentiation capacity. Biotechnol Appl Biochem 2007.

Chapter 9

General Discussion

Page 162

9.1 GENERAL DISCUSSION

Lacking vascular, neural, and lymphatic supply, articular cartilage is a comparatively inert tissue with no intrinsic capacity to repair following injury. Furthermore, the inability of neighbouring chondrocytes to taxi through their dense extracellular matrix into such defect voids and lay down neomatrix, translates to poor structural and functional outcomes in articular cartilage defects, and leads to degenerative changes within the joint. Hence these defects present a significant therapeutic dilemma to the orthopaedic surgeon. In 1965, Smith began a revolution in cell-based cartilage repair research when she isolated viable chondrocytes from articular cartilage and transplanted them into fresh cartilage nodules (Smith, 1965). Smith was unsuccessful, but with the advent of new technologies and improved techniques, the application of autologous chondrocyte implantation (ACI) has received much research attention and has subsequently been widely used in clinical orthopaedics (Brittberg et al., 1994; Grande et al., 1989; King et al., 2002; Peterson et al., 2002; Richardson et al., 1999; Roberts et al.,

2001; Yates, 2003). At present, matrix-induced autologous chondrocyte implantation

(MACI) is the most surgically simple form of ACI, and boasts clinical outcomes comparable to any technique on the market (Bartlett et al., 2005; Behrens et al., 2006;

Gigante et al., 2006; Marlovits et al., 2005). But whilst MACI has been extensively utilized for articular cartilage repair, many of the biological, histological, and clinical factors governing its success are still largely understudied.

Fibrin sealant and collagen membrane are key material components of the MACI technique, functioning both in the implantation and postoperative repair phases of the technology. Contrary to previous reports, we have shown that chondrocytes migrate into fibrin sealant (Tisseel®) from the collagen membrane (Brittberg et al., 1997; Kirilak et al., 2006). Moreover, thrombin – the major active ingredient in Tisseel®, was found to Page 163 stimulate chondrocyte proliferation, most likely through the protease-activated receptor-

1 (PAR-1) signaling pathway (Kirilak et al., 2006). This knowledge opens up the potential of thrombin and the PAR-1 pathway as therapeutic targets for improving chondrocyte migration and proliferation in the repairing defect site post-MACI. Given that Abiraman et al have also published results illustrating the osteoinductive capacity of fibrin sealant, this biomaterial may also be utilized in osteochondral defects

(Abiraman et al., 2002). Additionally, we have shown that the MACI type I/II collagen membrane maintains chondrocyte phenotype during the in vitro cultivation phase of the procedure, with a spherical morphology plus high S-100 and collagen II expression noted after passage (Zheng et al., 2007). This is an important finding as it has been shown by previous studies that chondrocytes cultured by traditional monolayer culture lose their phenotype (collagen II and aggrecan) after 2-3 passages (Zheng et al., 2004;

Gigante et al., 2007). Hence, although some surgeons still use the periosteal ACI technique, MACI is more biologically desirable because, as shown by the classic study of Benya and Shaffer, chondrocytes dedifferentiated by monolayer culture can be redifferentiated upon transfer to a 3D culture system such as collagen membrane (Benya and Shaffer, 1982). Regardless, it is likely that much of the observed qualitative repair variability in ACI treated patients is, at least in part, attributable to phenotypic instability introduced during the monolayer cultivation period. It is this phenotypic fluctuation, which has led to an increase of research into ACI technologies incorporating growth factors and biomechanical stimuli for the conditioning of chondrocytes prior to implantation.

We have validated the use of laser-scanning confocal arthroscopy (LSCA) for cartilage repair assessment following MACI in a sheep model, with significant intermodality agreement between this novel technology and traditional MRI and the ICRS repair Page 164 assessment methods (Jones et al., 2007). The non-destructive nature of LSCA provides a lower risk alternative to mechanical biopsy, but cannot reach the same depth for analysis. The technology loses resolution approximately 1mm into the tissue, limiting the application of LSCA to characterizing superficial tissue only. However, with an increased depth of resolute image acquisition, LSCA could prove a valuable tool for non-destructively characterizing cartilage repair ultrastructure. Whilst not very useful for full-thickness cartilage defects currently, LSCA may be applied to osteoarthritis research for studying surface changes early in the disease’s pathogenesis. This study also confirmed the commonly held notion that MACI-induced cartilage repair is qualitatively superior to native tissue repair. This is important because although there are local stem cells and some chondrocytes available to contribute to cartilage repair, they do not make a significant qualitative or quantitative contribution in most articular cartilage defects, especially when compared to MACI outcomes. Native repair is generally characterized by defects partially filled with fibrous tissue, fibrocartilage, or a mix of both.

In terms of repair histology, we have illustrated the progressive maturation of cartilage repair tissue from 21 days postoperatively, with hyaline-like cartilage observed at 6 months (Zheng et al., 2007). Given this repair timeline is akin to the timeline of functional restoration observed clinically in many patients, it would appear that the progression of structural and functional restoration go hand-in-hand; although this would be extremely difficult to confirm given the possible deleterious functional effects of biopsying a healthy repair site during graft maturation. Additionally, many surgeons may argue that the histological maturation of repair tissue is not important and is inevitably second to functional improvement. Whilst the improvement of patient function and quality of life is certainly the primary goal of treatments such as MACI, a Page 165 better understanding of repair biology may facilitate more sensitive postoperative care and rehabilitation protocols for optimizing functional improvement.

In terms of revised and clinically failed biopsies, we found both were mainly hyaline- like and fibrocartilage in repair type, and did not significantly differ in composition. As hyaline-like and fibrocartilage repair are also commonly observed in clinically successful MACI cases, as previously stated, it is feasible to speculate that repair histology bares no correlation to the patient’s clinical outcome. However, as the retrieval of biopsy tissue from both revised cases and cases with good to excellent clinical outcome is limited to a small area of the repair site, thereby not representative of the entire tissue, it is not possible to make definitive conclusions regarding any relationships between histology and other factors. In order to correlate histology to functional outcome, a significant cross-section of the repair site would need to be assessed; an action that would certainly compromise repair integrity and negate any research benefits. Future improvements to contrast-enhanced MRI or confocal arthroscopy technology may offer a non-invasive alternative to examining the holistic ultrastructure of these repair sites.

Further analysis of revised and failed repairs in terms of repair type, showed several differences in composition. Fibrocartilage was more Ki-67 positive compared to hyaline-like. As Ki-67 is a proliferative marker, this suggests that fibrocartilage has a higher comparative proliferative capacity to hyaline-like cartilage. Given that most fibrocartilage biopsies were taken from revised cases, this finding is consistent with the hypertrophic nature of tissue on the defect surface of revision cases, a region of excessive proliferation. Although notably, ACI using collagen membrane has been shown to have a lower incidence of graft hypertrophy than ACI using periosteum Page 166

(Gooding et al., 2006; Behrens et al., 2006). When looking at matrix, the improved content and localization of aggrecan, collagen II and collagen IX (matrix organizer) seen in hyaline-like cartilage repair (compared to fibrocartilage repair) is more consistent with the matrix characteristics of native hyaline cartilage. However, the pericellular-specific matrix localization of collagen VI, a key communicator of biomechanical stimuli seen in healthy hyaline-like articular cartilage, was lost in hyaline-like repairs, with diffuse staining observed instead. This disparity highlights an important point for all ACI technologies. While therapeutically, cell-based treatments such as MACI introduce stable chondrocytes with matrix protein expression profiles similar to that of the native tissue, the restoration of matrix composition must be accompanied by cellular ultrastructure if complete and functional regeneration is to occur. For example, conservation of the pericellular matrix is thought to be important for the conversion of biomechanical signals from the joint surface into chondrocyte remodeling, especially in times of rapid turnover such as repair (Morel and Quinn,

2004; Quinn et al., 1998). Similarly, the zonal architecture of healthy hyaline articular cartilage, lost in hyaline-like cartilage, in known to be specific to biomechanical transduction through the joint (Hasler et al., 1999). However, such complex ultrastructural and physiological properties of healthy hyaline cartilage are formed over years of development and cannot be easily replicated. Subsequently, the holistic regeneration of healthy hyaline articular cartilage following injury poses a sizeable challenge for those researching cell-based cartilage repair.

Although the structural repair of cartilage defects by MACI is an important component of therapeutic outcome, many surgeons and certainly most patients, will argue that joint function and quality of life is a more meaningful measure of success. To this end, we have shown herein that MACI produces significant improvements in both structural

(MRI) and functional outcome from 3 to 24 months postoperatively, with the strongest Page 167 gains evident in the first 12 months. These findings were consistent with previous findings, although interestingly we observed a drop in scoring at the 6 month time point the more strenuous parameters (sports, distance walking). The transient decreases were most likely attributable to the trauma of surgery, early postoperative restraints used to protect the graft, and graduated return to weightbearing (Hambly et al., 2006; Robertson et al., 2007). This illustrates the play-off between safely promoting graft repair and maturation, and reinstating the patient’s function and quality of life as soon as possible.

So whilst MACI may reduce patient function in the immediate postoperative period, over time, function is improved compared to presurgery and appears to be stable to at least 2 years. Consistent with previous studies, a significant correlation between MRI and functional outcome was also observed (Marlovits et al., 2006; Takahashi et al.,

2006). In saying this, while MRI has a place in the postoperative evaluation of MACI repair infill, integration, surface continuity, subchondral edema, and effusion, it is unlikely to ever be used as a prognostic indicator of functional improvement because of the aforementioned disparities between tissue structure and function.

Lastly, our retrospective patient survey has illustrated that the improvements in clinical outcome reported by the surgeon are shared by the patient’s sentiment, with encouraging results gained by retrospectively surveying 202 MACI patients for satisfaction. Eight-three percent of those surveyed reported a good or excellent overall outcome, as well as strong improvements noted in pain relief, return to sporting activity, swelling, and locking catching. Moreover, most patients rated their overall satisfaction with MACI as good or excellent, and would undergo the technique again if injured.

Together with our functional and MRI investigation, these results indicate that, from both the surgeon’s and the patient’s perspective, MACI is a reliable technique for significantly improving patient symptoms, function, and quality of life following Page 168 articular cartilage injury. Interestingly the survey also highlighted the influence of biomechanical stimulation of the cartilage on patient outcome, with significant improvements observed with elevated activity and weightbearing capacity. From the benchtop to the bedside, the biomechanical conditioning of chondrocytes is thought to translate into improved cartilage repair composition and knee function (Kreuz et al.,

2007; Marlovits et al., 2003; Mauck et al., 2003; Waldman et al., 2003). Notably, Kreuz et al recently reported significantly improved knee function in patients who regularly exercised preoperatively compared to those with rare or no sports involvement (Kreuz et al., 2007). Accordingly, and consistent with the literature, our research advocates the benefit of everyday exercise, postoperative rehabilitation, and return to sport following surgery, in order to maximise cartilage health and the efficacy of MACI® technology in cartilage repair.

Page 169

9.2 CONCLUSION

In summary, from the molecular aspects of the technique to clinical outcomes, this thesis has demonstrated a collection of evidence that suggests MACI as a credible cell- based treatment for articular cartilage defects. Our in vitro has supported the use of fibrin sealant and collagen membrane as the major components of MACI, illustrating improved chondrocyte proliferation, migration, and phenotypic stability. The potential of laser-scanning confocal arthroscopy (LSCA) for the non-destructive assessment of cartilage repair has been illustrated in a large animal model for the first time, yet improvements are needed to warrant clinical application. Histologically, we have evidenced the ability of MACI to stimulate hyaline-like cartilage repair by 6 months; that revised and failed MACI cases are a mixture of fibrocartilage and hyaline-like repair; and that hyaline-like repair is superior to fibrocartilage in terms of its matrix composition. Although, histological evaluation may not be representative of the whole repair site. And clinically, we have documented significant improvements in patient repair structure, function, symptoms, quality of life, and satisfaction, at the same time confirming sentiment within the literature regarding the importance of exercise/ rehabilitation for maximising MACI outcome. Concluding, the findings presented in this thesis suggest that matrix-induced autologous chondrocyte implantation (MACI) is a biologically justified and clinically efficacious cell-based treatment option for repairing articular cartilage defects.

Bibliography

Page 170

BIBLIOGRAPHY

Abiraman, S., H.K. Varma, P.R. Umashankar, and A. John. 2002. Fibrin glue as an osteoinductive protein in a mouse model. Biomaterials. 23:3023-31. Abraham, L.A., C. Chinni, A.L. Jenkins, A. Lourbakos, N. Ally, R.N. Pike, and E.J. Mackie. 2000. Expression of protease-activated receptor-2 by osteoblasts. Bone. 26:7-14. Abraham, L.A., and E.J. MacKie. 1999. Modulation of osteoblast-like cell behavior by activation of protease-activated receptor-1. J Bone Miner Res. 14:1320-9. Adachi, N., M. Ochi, Y. Uchio, J. Iwasa, S. Furukawa, and M. Deie. 2004. Osteochondral lesion located at the lateral femoral condyle reconstructed by the transplantation of tissue-engineered cartilage in combination with a periosteum with bone block: a case report. Knee Surg Sports Traumatol Arthrosc. 12:444-7. Agneskirchner, J.D., P. Brucker, A. Burkart, and A.B. Imhoff. 2002. Large osteochondral defects of the femoral condyle: press-fit transplantation of the posterior femoral condyle (MEGA-OATS). Knee Surg Sports Traumatol Arthrosc. 10:160-8. Akbar, A., V. Diedrichs, and U.A. Wagner. 2003. [Massive arthrofibrosis of the knee joint after carbon fibre rod implantation. Two cases]. Z Orthop Ihre Grenzgeb. 141:570-2. Akers, I.A., M. Parsons, M.R. Hill, M.D. Hollenberg, S. Sanjar, G.J. Laurent, and R.J. McAnulty. 2000. Mast cell tryptase stimulates human lung fibroblast proliferation via protease-activated receptor-2. Am J Physiol Lung Cell Mol Physiol. 278:L193-201. Akeson, W.H., W. Bugbee, C. Chu, and A. Giurea. 2001. Differences in mesenchymal tissue repair. Clin Orthop Relat Res:S124-41. Albala, D.M. 2003. Fibrin sealants in clinical practice. Cardiovasc Surg. 11 Suppl 1:5- 11. Allen, R.T., J.P. Tasto, J. Cummings, C.M. Robertson, and D. Amiel. 2006. Meniscal debridement with an arthroscopic radiofrequency wand versus an arthroscopic shaver: comparative effects on menisci and underlying articular cartilage. Arthroscopy. 22:385-93. Alleyne, K.R., and M.T. Galloway. 2001. Management of osteochondral injuries of the knee. Clin Sports Med. 20:343-64. Amin, A.A., W. Bartlett, C.R. Gooding, M. Sood, J.A. Skinner, R.W. Carrington, T.W. Briggs, and G. Bentley. 2005. The use of autologous chondrocyte implantation

Page 171

following and combined with anterior cruciate ligament reconstruction. Int Orthop:1-6. Andereya, S., U. Maus, K. Gavenis, R. Muller-Rath, O. Miltner, T. Mumme, and U. Schneider. 2006. [First clinical experiences with a novel 3D-collagen gel (CaReS) for the treatment of focal cartilage defects in the knee]. Z Orthop Ihre Grenzgeb. 144:272-80. Armstrong, C.G., W.M. Lai, and V.C. Mow. 1984. An analysis of the unconfined compression of articular cartilage. J Biomech Eng. 106:165-73. Aroen, A., S. Loken, S. Heir, E. Alvik, A. Ekeland, O.G. Granlund, and L. Engebretsen. 2004. Articular cartilage lesions in 993 consecutive knee arthroscopies. Am J Sports Med. 32:211-5. Asokananthan, N., P.T. Graham, J. Fink, D.A. Knight, A.J. Bakker, A.S. McWilliam, P.J. Thompson, and G.A. Stewart. 2002. Activation of protease-activated receptor (PAR)-1, PAR-2, and PAR-4 stimulates IL-6, IL-8, and prostaglandin E2 release from human respiratory epithelial cells. J Immunol. 168:3577-85. Bachmann, G., E. Basad, D. Lommel, and J. Steinmeyer. 2004. [MRI in the follow-up of matrix-supported autologous chondrocyte transplantation (MACI) and microfracture]. Radiologe. 44:773-82. Bar-Shavit, R., A. Kahn, J.W. Fenton, 2nd, and G.D. Wilner. 1983a. Chemotactic response of monocytes to thrombin. J Cell Biol. 96:282-5. Bar-Shavit, Z., S.L. Teitelbaum, P. Reitsma, A. Hall, L.E. Pegg, J. Trial, and A.J. Kahn. 1983b. Induction of monocytic differentiation and bone resorption by 1,25- dihydroxyvitamin D3. Proc Natl Acad Sci U S A. 80:5907-11. Bartlett, W., C.R. Gooding, R.W. Carrington, J.A. Skinner, T.W. Briggs, and G. Bentley. 2005a. Autologous chondrocyte implantation at the knee using a bilayer collagen membrane with bone graft. A preliminary report. J Bone Joint Surg Br. 87:330-2. Bartlett, W., J.A. Skinner, C.R. Gooding, R.W. Carrington, A.M. Flanagan, T.W. Briggs, and G. Bentley. 2005b. Autologous chondrocyte implantation versus matrix-induced autologous chondrocyte implantation for osteochondral defects of the knee: a prospective, randomised study. J Bone Joint Surg Br. 87:640-5. Basad, E., H. Stürz, and J. Steinmeyer. 2004. Treatment of chondral defects with MACI or microfracture. First results of a comparative clinical study [Die behandlung chondraler defekte mit MACI oder microfracture - erste Ergebnisse einer vergleichenden klinischen Studie]. Orthopädische Praxis. 40:6-10.

Page 172

Behrens, P., T. Bitter, B. Kurz, and M. Russlies. 2006. Matrix-associated autologous chondrocyte transplantation/implantation (MACT/MACI)--5-year follow-up. Knee. 13:194-202. Bentley, G., L.C. Biant, R.W. Carrington, M. Akmal, A. Goldberg, A.M. Williams, J.A. Skinner, and J. Pringle. 2003. A prospective, randomised comparison of autologous chondrocyte implantation versus mosaicplasty for osteochondral defects in the knee. J Bone Joint Surg Br. 85:223-30. Bentley, G., and T. Minas. 2000. Treating joint damage in young people. Bmj. 320:1585-8. Benya, P.D., and J.D. Shaffer. 1982. Dedifferentiated chondrocytes reexpress the differentiated collagen phenotype when cultured in agarose gels. Cell. 30:215- 24. Bi, W., J.M. Deng, Z. Zhang, R.R. Behringer, and B. de Crombrugghe. 1999. Sox9 is required for cartilage formation. Nat Genet. 22:85-9. Birk, G.T., and J.C. DeLee. 2001. Osteochondral injuries. Clinical findings. Clin Sports Med. 20:279-86. Blackhart, B.D., K. Emilsson, D. Nguyen, W. Teng, A.J. Martelli, S. Nystedt, J. Sundelin, and R.M. Scarborough. 1996. Ligand cross-reactivity within the protease-activated receptor family. J Biol Chem. 271:16466-71. Bobic, V. 1996. Arthroscopic osteochondral autograft transplantation in anterior cruciate ligament reconstruction: a preliminary clinical study. Knee Surg Sports Traumatol Arthrosc. 3:262-4. Boden, B.P., A.W. Pearsall, W.E. Garrett, Jr., and J.A. Feagin, Jr. 1997. Patellofemoral Instability: Evaluation and Management. J Am Acad Orthop Surg. 5:47-57. Bonyadi, M., S.D. Waldman, D. Liu, J.E. Aubin, M.D. Grynpas, and W.L. Stanford. 2003. Mesenchymal progenitor self-renewal deficiency leads to age-dependent osteoporosis in Sca-1/Ly-6A null mice. Proc Natl Acad Sci U S A. 100:5840-5. Borrelli, V., A.V. Sterpetti, P. Coluccia, B. Randone, A. Cavallaro, L. Santoro D'Angelo, and A. Cucina. 2001. Bimodal concentration-dependent effect of thrombin on endothelial cell proliferation and growth factor release in culture. J Surg Res. 100:154-60. Brass, L.F., and M. Molino. 1997. Protease-activated G protein-coupled receptors on human platelets and endothelial cells. Thromb Haemost. 78:234-41.

Page 173

Breinan, H.A., T. Minas, H.P. Hsu, S. Nehrer, S. Shortkroff, and M. Spector. 2001. Autologous chondrocyte implantation in a canine model: change in composition of reparative tissue with time. J Orthop Res. 19:482-92. Briggs, T.W., S. Mahroof, L.A. David, J. Flannelly, J. Pringle, and M. Bayliss. 2003. Histological evaluation of chondral defects after autologous chondrocyte implantation of the knee. J Bone Joint Surg Br. 85:1077-83. Brittberg, M., E. Faxen, and L. Peterson. 1994a. Carbon fiber scaffolds in the treatment of early knee osteoarthritis. A prospective 4-year followup of 37 patients. Clin Orthop Relat Res:155-64. Brittberg, M., A. Lindahl, A. Nilsson, C. Ohlsson, O. Isaksson, and L. Peterson. 1994b. Treatment of deep cartilage defects in the knee with autologous chondrocyte transplantation. N Engl J Med. 331:889-95. Brittberg, M., A. Nilsson, A. Lindahl, C. Ohlsson, and L. Peterson. 1996. Rabbit articular cartilage defects treated with autologous cultured chondrocytes. Clin Orthop Relat Res:270-83. Brittberg, M., E. Sjogren-Jansson, A. Lindahl, and L. Peterson. 1997. Influence of fibrin sealant (Tisseel) on osteochondral defect repair in the rabbit knee. Biomaterials. 18:235-42. Brittberg, M., and C.S. Winalski. 2003. Evaluation of cartilage injuries and repair. J Bone Joint Surg Am. 85-A Suppl 2:58-69. Brooks, P.M. 2006. The burden of musculoskeletal disease--a global perspective. Clin Rheumatol. 25:778-81. Brown, L.F., N. Lanir, J. McDonagh, K. Tognazzi, A.M. Dvorak, and H.F. Dvorak. 1993. Fibroblast migration in fibrin gel matrices. Am J Pathol. 142:273-83. Browne, J.E., A.F. Anderson, R. Arciero, B. Mandelbaum, J.B. Moseley, Jr., L.J. Micheli, F. Fu, and C. Erggelet. 2005. Clinical outcome of autologous chondrocyte implantation at 5 years in US subjects. Clin Orthop Relat Res:237- 45. Bugbee, W.D., and F.R. Convery. 1999. Osteochondral allograft transplantation. Clin Sports Med. 18:67-75. Burton-Wurster, N., R.G. Mateescu, R.J. Todhunter, K.M. Clements, Q. Sun, V. Scarpino, and G. Lust. 2005. Genes in canine articular cartilage that respond to mechanical injury: gene expression studies with Affymetrix canine GeneChip. J Hered. 96:821-8.

Page 174

Burton-Wurster, N., M. Vernier-Singer, T. Farquhar, and G. Lust. 1993. Effect of compressive loading and unloading on the synthesis of total protein, proteoglycan, and fibronectin by canine cartilage explants. J Orthop Res. 11:717-29. Buschmann, M.D., J. Soulhat, A. Shirazi-Adl, J.S. Jurvelin, and E.B. Hunziker. 1998. Confined compression of articular cartilage: linearity in ramp and sinusoidal tests and the importance of interdigitation and incomplete confinement. J Biomech. 31:171-8. Canonico, S. 2003. The use of human fibrin glue in the surgical operations. Acta Biomed Ateneo Parmense. 74 Suppl 2:21-5. Chaipinyo, K., B.W. Oakes, and M.P. Van Damme. 2004. The use of debrided human articular cartilage for autologous chondrocyte implantation: maintenance of chondrocyte differentiation and proliferation in type I collagen gels. J Orthop Res. 22:446-55. Chang, R.W., J. Falconer, S.D. Stulberg, W.J. Arnold, L.M. Manheim, and A.R. Dyer. 1993. A randomized, controlled trial of arthroscopic surgery versus closed- needle joint lavage for patients with osteoarthritis of the knee. Arthritis Rheum. 36:289-96. Charriere, G., M. Bejot, L. Schnitzler, G. Ville, and D.J. Hartmann. 1989. Reactions to a bovine collagen implant. Clinical and immunologic study in 705 patients. J Am Acad Dermatol. 21:1203-8. Cherubino, P., F.A. Grassi, P. Bulgheroni, and M. Ronga. 2003. Autologous chondrocyte implantation using a bilayer collagen membrane: a preliminary report. J Orthop Surg (Hong Kong). 11:10-5. Chiang, H., T.F. Kuo, C.C. Tsai, M.C. Lin, B.R. She, Y.Y. Huang, H.S. Lee, C.S. Shieh, M.H. Chen, J.A. Ramshaw, J.A. Werkmeister, R.S. Tuan, and C.C. Jiang. 2005. Repair of porcine articular cartilage defect with autologous chondrocyte transplantation. J Orthop Res. 23:584-93. Chinni, C., M.R. de Niese, A.L. Jenkins, R.N. Pike, S.P. Bottomley, and E.J. Mackie. 2000. Protease-activated receptor-2 mediates proliferative responses in skeletal myoblasts. J Cell Sci. 113 Pt 24:4427-33. Chinni, C., M.R. de Niese, D.J. Tew, A.L. Jenkins, S.P. Bottomley, and E.J. Mackie. 1999. Thrombin, a survival factor for cultured myoblasts. J Biol Chem. 274:9169-74.

Page 175

Chu, C.R., J.S. Dounchis, M. Yoshioka, R.L. Sah, R.D. Coutts, and D. Amiel. 1997. Osteochondral repair using perichondrial cells. A 1-year study in rabbits. Clin Orthop Relat Res:220-9. Chung, C.B., L.R. Frank, and D. Resnick. 2001. Cartilage imaging techniques: current clinical applications and state of the art imaging. Clin Orthop Relat Res:S370-8. Ciano, P.S., R.B. Colvin, A.M. Dvorak, J. McDonagh, and H.F. Dvorak. 1986. Macrophage migration in fibrin gel matrices. Lab Invest. 54:62-70. Clanton, T.O., and J.C. DeLee. 1982. Osteochondritis dissecans. History, pathophysiology and current treatment concepts. Clin Orthop Relat Res:50-64. Clar, C., E. Cummins, L. McIntyre, S. Thomas, J. Lamb, L. Bain, P. Jobanputra, and N. Waugh. 2005. Clinical and cost-effectiveness of autologous chondrocyte implantation for cartilage defects in knee joints: systematic review and economic evaluation. Health Technol Assess. 9:iii-iv, ix-x, 1-82. Convery, F.R., M.H. Meyers, and W.H. Akeson. 1991. Fresh osteochondral allografting of the femoral condyle. Clin Orthop Relat Res:139-45. Corvera, C.U., O. Dery, K. McConalogue, P. Gamp, M. Thoma, B. Al-Ani, G.H. Caughey, M.D. Hollenberg, and N.W. Bunnett. 1999. Thrombin and mast cell tryptase regulate guinea-pig myenteric neurons through proteinase-activated receptors-1 and -2. J Physiol. 517 ( Pt 3):741-56. Curl, W.W., J. Krome, E.S. Gordon, J. Rushing, B.P. Smith, and G.G. Poehling. 1997. Cartilage injuries: a review of 31,516 knee arthroscopies. Arthroscopy. 13:456- 60. Currie, L.J., J.R. Sharpe, and R. Martin. 2001. The use of fibrin glue in skin grafts and tissue-engineered skin replacements: a review. Plast Reconstr Surg. 108:1713- 26. Curtin, W., W. Reville, M. Heapes, J. Lyons, and D. Muckle. 1994. The chondrogenic potential of carbon fiber and carbon fiber periosteum implants: an ultrastructural study in the rabbit. Osteoarthritis Cartilage. 2:253-8. De Bari, C., F. Dell'Accio, and F.P. Luyten. 2004. Failure of in vitro-differentiated mesenchymal stem cells from the synovial membrane to form ectopic stable cartilage in vivo. Arthritis Rheum. 50:142-50. de Niese, M.R., C. Chinni, R.N. Pike, S.P. Bottomley, and E.J. Mackie. 2002. Dissection of protease-activated receptor-1-dependent and -independent responses to thrombin in skeletal myoblasts. Exp Cell Res. 274:149-56.

Page 176

Debnath, U.K., J.A. Fairclough, and R.L. Williams. 2004. Long-term local effects of carbon fibre in the knee. Knee. 11:259-64. Delaney, P.M., M.R. Harris, and R.G. King. 1993. Novel microscopy using fibre optic confocal imaging and its suitability for subsurface blood vessel imaging in vivo. Clin Exp Pharmacol Physiol. 20:197-8. Dell'Accio, F., J. Vanlauwe, J. Bellemans, J. Neys, C. De Bari, and F.P. Luyten. 2003. Expanded phenotypically stable chondrocytes persist in the repair tissue and contribute to cartilage matrix formation and structural integration in a goat model of autologous chondrocyte implantation. J Orthop Res. 21:123-31. Derrett, S., E.A. Stokes, M. James, W. Bartlett, and G. Bentley. 2005. Cost and health status analysis after autologous chondrocyte implantation and mosaicplasty: a retrospective comparison. Int J Technol Assess Health Care. 21:359-67. Dery, O., C.U. Corvera, M. Steinhoff, and N.W. Bunnett. 1998. Proteinase-activated receptors: novel mechanisms of signaling by serine proteases. Am J Physiol. 274:C1429-52. Dorotka, R., U. Windberger, K. Macfelda, U. Bindreiter, C. Toma, and S. Nehrer. 2005. Repair of articular cartilage defects treated by microfracture and a three- dimensional collagen matrix. Biomaterials. 26:3617-29. Dowthwaite, G.P., J.C. Bishop, S.N. Redman, I.M. Khan, P. Rooney, D.J. Evans, L. Haughton, Z. Bayram, S. Boyer, B. Thomson, M.S. Wolfe, and C.W. Archer. 2004. The surface of articular cartilage contains a progenitor cell population. J Cell Sci. 117:889-97. Dozin, B., M. Malpeli, R. Cancedda, P. Bruzzi, S. Calcagno, L. Molfetta, F. Priano, E. Kon, and M. Marcacci. 2005. Comparative evaluation of autologous chondrocyte implantation and mosaicplasty: a multicentered randomized clinical trial. Clin J Sport Med. 15:220-6. Dragoo, J.L., B. Samimi, M. Zhu, S.L. Hame, B.J. Thomas, J.R. Lieberman, M.H. Hedrick, and P. Benhaim. 2003. Tissue-engineered cartilage and bone using stem cells from human infrapatellar fat pads. J Bone Joint Surg Br. 85:740-7. Driesang, I.M., and E.B. Hunziker. 2000. Delamination rates of tissue flaps used in articular cartilage repair. J Orthop Res. 18:909-11. Edwards, R.B., 3rd, Y. Lu, S. Nho, B.J. Cole, and M.D. Markel. 2002a. Thermal chondroplasty of chondromalacic human cartilage. An ex vivo comparison of bipolar and monopolar radiofrequency devices. Am J Sports Med. 30:90-7.

Page 177

Edwards, R.B., 3rd, Y. Lu, E. Rodriguez, and M.D. Markel. 2002b. Thermometric determination of cartilage matrix temperatures during thermal chondroplasty: comparison of bipolar and monopolar radiofrequency devices. Arthroscopy. 18:339-46. Ehlers, E.M., P. Behrens, L. Wunsch, W. Kuhnel, and M. Russlies. 2001. Effects of hyaluronic acid on the morphology and proliferation of human chondrocytes in primary cell culture. Ann Anat. 183:13-7. Elders, M.J. 2000. The increasing impact of arthritis on public health. J Rheumatol Suppl. 60:6-8. Enright, P.L. 2003. The six-minute walk test. Respir Care. 48:783-5. Evans, P.J., A. Miniaci, and M.B. Hurtig. 2004. Manual punch versus power harvesting of osteochondral grafts. Arthroscopy. 20:306-10. Facchini, A., G. Lisignoli, S. Cristino, L. Roseti, L. De Franceschi, E. Marconi, and B. Grigolo. 2006. Human chondrocytes and mesenchymal stem cells grown onto engineered scaffold. Biorheology. 43:471-80. Farmer, J.M., D.F. Martin, C.A. Boles, and W.W. Curl. 2001. Chondral and osteochondral injuries. Diagnosis and management. Clin Sports Med. 20:299- 320. Feczko, P., L. Hangody, J. Varga, L. Bartha, Z. Dioszegi, G. Bodo, Z. Kendik, and L. Modis. 2003. Experimental results of donor site filling for autologous osteochondral mosaicplasty. Arthroscopy. 19:755-61. Feder, J., H.D. Adkisson, N. Kizer, and e. al. 2004. The promise of chondral repair using neocartilage. In Tissue Engineering in Musculoskeletal Clinical Practice. L. Sandell and A. Grodzinsky, editors. American Academy of Orthopaedic Surgeons, Rosemont. 219-226. Fitzgerald, J.B., M. Jin, D. Dean, D.J. Wood, M.H. Zheng, and A.J. Grodzinsky. 2004. Mechanical compression of cartilage explants induces multiple time-dependent gene expression patterns and involves intracellular calcium and cyclic AMP. J Biol Chem. 279:19502-11. Fraser, S.A., A. Crawford, A. Frazer, S. Dickinson, A.P. Hollander, I.M. Brook, and P.V. Hatton. 2006. Localization of Type VI Collagen in Tissue-Engineered Cartilage on Polymer Scaffolds. Tissue Eng. Frenkel, S.R., B. Toolan, D. Menche, M.I. Pitman, and J.M. Pachence. 1997. Chondrocyte transplantation using a collagen bilayer matrix for cartilage repair. J Bone Joint Surg Br. 79:831-6.

Page 178

Frisbie, D.D., J.T. Oxford, L. Southwood, G.W. Trotter, W.G. Rodkey, J.R. Steadman, J.L. Goodnight, and C.W. McIlwraith. 2003. Early events in cartilage repair after subchondral bone microfracture. Clin Orthop Relat Res:215-27. Frost, A., K.B. Jonsson, P. Ridefelt, O. Nilsson, S. Ljunghall, and O. Ljunggren. 1999. Thrombin, but not bradykinin, stimulates proliferation in isolated human osteoblasts, via a mechanism not dependent on endogenous prostaglandin formation. Acta Orthop Scand. 70:497-503. Fu, F.H. 2005. Rate of improvement was not different after osteochondral repair with matrix-induced autologous chondrocyte implantation or autologous chondrocyte implantation with a cover made from porcine-derived type I/type III collagen. J Bone Joint Surg Am. 87:2593. Fu, F.H., D. Zurakowski, J.E. Browne, B. Mandelbaum, C. Erggelet, J.B. Moseley, Jr., A.F. Anderson, and L.J. Micheli. 2005. Autologous chondrocyte implantation versus debridement for treatment of full-thickness chondral defects of the knee: an observational cohort study with 3-year follow-up. Am J Sports Med. 33:1658- 66. Furthmayr, H., and R. Timpl. 1976. Immunochemistry of collagens and procollagens. Int Rev Connect Tissue Res. 7:61-99. Fuss, M., E.M. Ehlers, M. Russlies, J. Rohwedel, and P. Behrens. 2000. Characteristics of human chondrocytes, osteoblasts and fibroblasts seeded onto a type I/III collagen sponge under different culture conditions. A light, scanning and transmission electron microscopy study. Ann Anat. 182:303-10. Galois, L., S. Hutasse, D. Cortial, C.F. Rousseau, L. Grossin, M.C. Ronziere, D. Herbage, and A.M. Freyria. 2006. Bovine chondrocyte behaviour in three- dimensional type I collagen gel in terms of gel contraction, proliferation and gene expression. Biomaterials. 27:79-90. Gigante, A., C. Bevilacqua, A. Ricevuto, M. Mattioli-Belmonte, and F. Greco. 2006. Membrane-seeded autologous chondrocytes: cell viability and characterization at surgery. Knee Surg Sports Traumatol Arthrosc. Gille, J., U. Meisner, E.M. Ehlers, A. Muller, M. Russlies, and P. Behrens. 2005. Migration pattern, morphology and viability of cells suspended in or sealed with fibrin glue: A histomorphologic study. Tissue Cell. 37:339-48. Gillogly, S.D. 2003. Treatment of large full-thickness chondral defects of the knee with autologous chondrocyte implantation. Arthroscopy. 19 Suppl 1:147-53.

Page 179

Gobbi, A., R.A. Francisco, J.H. Lubowitz, F. Allegra, and G. Canata. 2006a. Osteochondral lesions of the talus: randomized controlled trial comparing chondroplasty, microfracture, and osteochondral autograft transplantation. Arthroscopy. 22:1085-92. Gobbi, A., E. Kon, M. Berruto, R. Francisco, G. Filardo, and M. Marcacci. 2006b. Patellofemoral full-thickness chondral defects treated with Hyalograft-C: a clinical, arthroscopic, and histologic review. Am J Sports Med. 34:1763-73. Gooding, C.R., W. Bartlett, G. Bentley, J.A. Skinner, R. Carrington, and A. Flanagan. 2006a. A prospective, randomised study comparing two techniques of autologous chondrocyte implantation for osteochondral defects in the knee: Periosteum covered versus type I/III collagen covered. Knee. 13:203-10. Gooding, C.R., W. Bartlett, G. Bentley, J.A. Skinner, R. Carrington, and A. Flanagan. 2006b. A prospective, ranomised study comparing two techniques of autologous chondrocyte implantation for osteochondral defects in the knee: Periosteum covered versus type I/III collagen covered. Knee. Grande, D.A., J. Mason, E. Light, and D. Dines. 2003. Stem cells as platforms for delivery of genes to enhance cartilage repair. J Bone Joint Surg Am. 85-A Suppl 2:111-6. Grande, D.A., M.I. Pitman, L. Peterson, D. Menche, and M. Klein. 1989. The repair of experimentally produced defects in rabbit articular cartilage by autologous chondrocyte transplantation. J Orthop Res. 7:208-18. Green, L.M., J.S. King, B.M. Bianski, M.M. Pink, and C.M. Jobe. 2006. In vitro effects of 3 common arthroscopic instruments on articular cartilage. Arthroscopy. 22:300-7. Grigolo, B., L. De Franceschi, L. Roseti, L. Cattini, and A. Facchini. 2005a. Down regulation of degenerative cartilage molecules in chondrocytes grown on a hyaluronan-based scaffold. Biomaterials. 26:5668-76. Grigolo, B., L. Roseti, L. De Franceschi, A. Piacentini, L. Cattini, M. Manfredini, R. Faccini, and A. Facchini. 2005b. Molecular and immunohistological characterization of human cartilage two years following autologous cell transplantation. J Bone Joint Surg Am. 87:46-57. Grigolo, B., L. Roseti, M. Fiorini, M. Fini, G. Giavaresi, N.N. Aldini, R. Giardino, and A. Facchini. 2001. Transplantation of chondrocytes seeded on a hyaluronan derivative (hyaff-11) into cartilage defects in rabbits. Biomaterials. 22:2417-24.

Page 180

Gruber, R., C. Jindra, B. Kandler, G. Watzak, M.B. Fischer, and G. Watzek. 2004. Proliferation of dental pulp fibroblasts in response to thrombin involves mitogen-activated protein kinase signalling. Int Endod J. 37:145-50. Gudas, R., E. Stankevicius, E. Monastyreckiene, D. Pranys, and R.J. Kalesinskas. 2006. Osteochondral autologous transplantation versus microfracture for the treatment of articular cartilage defects in the knee joint in athletes. Knee Surg Sports Traumatol Arthrosc. Haddo, O., S. Mahroof, D. Higgs, L. David, J. Pringle, M. Bayliss, S.R. Cannon, and T.W. Briggs. 2004. The use of chondrogide membrane in autologous chondrocyte implantation. Knee. 11:51-5. Hafez, M.I., R.R. Coombs, S. Zhou, and I.D. McCarthy. 2002. Ablation of bone, cartilage, and facet joint capsule using Ho:YAG laser. J Clin Laser Med Surg. 20:251-5. Hambly, K., V. Bobic, B. Wondrasch, D. Van Assche, and S. Marlovits. 2006. Autologous Chondrocyte Implantation Postoperative Care and Rehabilitation: Science and Practice. Am J Sports Med. Hangody, L., and P. Fules. 2003. Autologous osteochondral mosaicplasty for the treatment of full-thickness defects of weight-bearing joints: ten years of experimental and clinical experience. J Bone Joint Surg Am. 85-A Suppl 2:25- 32. Hangody, L., G. Kish, Z. Karpati, I. Szerb, and I. Udvarhelyi. 1997. Arthroscopic autogenous osteochondral mosaicplasty for the treatment of femoral condylar articular defects. A preliminary report. Knee Surg Sports Traumatol Arthrosc. 5:262-7. Hasler, E.M., W. Herzog, J.Z. Wu, W. Muller, and U. Wyss. 1999. Articular cartilage biomechanics: theoretical models, material properties, and biosynthetic response. Crit Rev Biomed Eng. 27:415-88. Hayes, D.W., Jr., R.L. Brower, and K.J. John. 2001. Articular cartilage. Anatomy, injury, and repair. Clin Podiatr Med Surg. 18:35-53. Henderson, I., B. Tuy, and B. Oakes. 2004. Reoperation after autologous chondrocyte implantation. Indications and findings. J Bone Joint Surg Br. 86:205-11. Henderson, I.J., and P. Lavigne. 2006. Periosteal autologous chondrocyte implantation for patellar chondral defect in patients with normal and abnormal patellar tracking. Knee. 13:274-9.

Page 181

Henderson, I.J., B. Tuy, D. Connell, B. Oakes, and W.H. Hettwer. 2003. Prospective clinical study of autologous chondrocyte implantation and correlation with MRI at three and 12 months. J Bone Joint Surg Br. 85:1060-6. Hendrickson, D.A., A.J. Nixon, H.N. Erb, and G. Lust. 1994a. Phenotype and biological activity of neonatal equine chondrocytes cultured in a three-dimensional fibrin matrix. Am J Vet Res. 55:410-4. Hendrickson, D.A., A.J. Nixon, D.A. Grande, R.J. Todhunter, R.M. Minor, H. Erb, and G. Lust. 1994b. Chondrocyte-fibrin matrix transplants for resurfacing extensive articular cartilage defects. J Orthop Res. 12:485-97. Hollander, A.P., S.C. Dickinson, J.S. T, P. Brun, R. Cortivo, E. Kon, M. Marcacci, S. Zanasi, A. Borrione, C.D. Luca, A. Pavesio, C. Soranzo, and G. Abatangelo. 2006. Maturation of Tissue Engineered Cartilage Implanted in Injured and Osteoarthritic Human Knees. Tissue Eng. Homminga, G.N., S.K. Bulstra, P.S. Bouwmeester, and A.J. van der Linden. 1990. Perichondral grafting for cartilage lesions of the knee. J Bone Joint Surg Br. 72:1003-7. Homminga, G.N., P. Buma, H.W. Koot, P.M. van der Kraan, and W.B. van den Berg. 1993. Chondrocyte behavior in fibrin glue in vitro. Acta Orthop Scand. 64:441- 5. Hoser, C., O. Bichler, R. Bale, R. Rosenberger, M. Rieger, P. Kovacs, T. Lang, and C. Fink. 2004. A computer assisted surgical technique for retrograde autologous osteochondral grafting in talar osteochondritis dissecans (OCD): a cadaveric study. Knee Surg Sports Traumatol Arthrosc. 12:65-71. Huang, F.S., P.T. Simonian, A.G. Norman, and J.M. Clark. 2004. Effects of small incongruities in a sheep model of osteochondral autografting. Am J Sports Med. 32:1842-8. Huang, L., J. Xu, D.J. Wood, and M.H. Zheng. 2000. Gene expression of osteoprotegerin ligand, osteoprotegerin, and receptor activator of NF-kappaB in giant cell tumor of bone: possible involvement in tumor cell-induced osteoclast- like cell formation. Am J Pathol. 156:761-7. Huntley, J.S., P.G. Bush, J.M. McBirnie, A.H. Simpson, and A.C. Hall. 2005. Chondrocyte death associated with human femoral osteochondral harvest as performed for mosaicplasty. J Bone Joint Surg Am. 87:351-60.

Page 182

Hunziker, E.B., and L.C. Rosenberg. 1996. Repair of partial-thickness defects in articular cartilage: cell recruitment from the synovial membrane. J Bone Joint Surg Am. 78:721-33. Hutmacher, D.W., J.C. Goh, and S.H. Teoh. 2001. An introduction to biodegradable materials for tissue engineering applications. Ann Acad Med Singapore. 30:183- 91. Im, G.I., N.H. Jung, and S.K. Tae. 2006. Chondrogenic differentiation of mesenchymal stem cells isolated from patients in late adulthood: the optimal conditions of growth factors. Tissue Eng. 12:527-36. Ishihara, H., A.J. Connolly, D. Zeng, M.L. Kahn, Y.W. Zheng, C. Timmons, T. Tram, and S.R. Coughlin. 1997. Protease-activated receptor 3 is a second thrombin receptor in humans. Nature. 386:502-6. Iwasa, J., M. Ochi, Y. Uchio, K. Katsube, N. Adachi, and K. Kawasaki. 2003. Effects of cell density on proliferation and matrix synthesis of chondrocytes embedded in atelocollagen gel. Artif Organs. 27:249-55. Jackson, M.R. 2001. Fibrin sealants in surgical practice: An overview. Am J Surg. 182:1S-7S. Jackson, R. 1991. Arthroscopic treatment of degenrative arthritis. Raven Press, New York. 319-23 pp. Jackson, R.W., H.J. Marans, and R.S. Silver. 1988. The arthroscopic treatment of degenerative arthritis of the knee. J Bone Joint Surg Br. 70:332. Jakobsen, R.B., L. Engebretsen, and J.R. Slauterbeck. 2005. An analysis of the quality of cartilage repair studies. J Bone Joint Surg Am. 87:2232-9. James, S.L., D.A. Connell, A. Saifuddin, J.A. Skinner, and T.W. Briggs. 2006. MR imaging of autologous chondrocyte implantation of the knee. Eur Radiol:1-9. Janecki, C.J., M.W. Perry, A.O. Bonati, and M. Bendel. 1998. Safe parameters for laser chondroplasty of the knee. Lasers Surg Med. 23:141-50. Jazrawi, L.M., A. Chen, D. Stein, C.S. Heywood, A. Bernstein, G. Steiner, and A. Rokito. 2003. The effects of radiofrequency bipolar thermal energy on human meniscal tissue. Bull Hosp Jt Dis. 61:114-7. Jenkins, A.L., M.D. Bootman, M.J. Berridge, and S.R. Stone. 1994. Differences in intracellular calcium signaling after activation of the thrombin receptor by thrombin and agonist peptide in osteoblast-like cells. J Biol Chem. 269:17104- 10.

Page 183

Jenkins, A.L., G.L. Howells, E. Scott, B.F. Le Bonniec, M.A. Curtis, and S.R. Stone. 1995. The response to thrombin of human neutrophils: evidence for two novel receptors. J Cell Sci. 108 ( Pt 9):3059-66. Johnson, L. 1991. Arthroscopic abrasion arthroplasty. Raven Press, New York. 341-60 pp. Jones, C.W., A. Keogh, D. Smolinski, J.P. Wu, T.B. Kirk, and M.H. Zheng. 2004a. Histological Assessment of the Chondral and Connective Tissues of the Knee by Confocal Arthroscope. Journal of Musculoskeletal Research. 8:75-86. Jones, C.W., D. Smolinski, J. Wu, and e. al. 2004b. Quantification of chondrocyte morphology by confocal arthroscopy. J Musculoskel Res. 8. Jones, C.W., C. Willers, A. Keogh, D. Smolinski, D. Fick, P.J. Yates, T.B. Kirk, and M.H. Zheng. 2007. Matrix-induced autologous chondrocyte implantation in sheep: objective assessments including confocal arthroscopy. J Orthop Res. Jurvelin, J.S., M.D. Buschmann, and E.B. Hunziker. 2003. Mechanical anisotropy of the human knee articular cartilage in compression. Proc Inst Mech Eng [H]. 217:215-9. Kaplonyi, G., I. Zimmerman, A.D. Frenyo, T. Farkas, and G. Nemes. 1988. The use of fibrin adhesive in the repair of chondral and osteochondral injuries. Injury. 19:267-72. Kaplonyi, G., I. Zimmermann, T. Farkas, and T. Viola. 1984. [Repair of cartilage injuries and osteochondral fractures with fibrin glue]. Orv Hetil. 125:2237-40, 2243. Karataglis, D., and D.J. Learmonth. 2005. Management of big osteochondral defects of the knee using osteochondral allografts with the MEGA-OATS technique. Knee. 12:389-93. Karp, J.M., T.S. Tanaka, R. Zohar, J. Sodek, M.S. Shoichet, J.E. Davies, and W.L. Stanford. 2005. Thrombin mediated migration of osteogenic cells. Bone. 37:337- 348. Katsube, K., M. Ochi, Y. Uchio, S. Maniwa, M. Matsusaki, M. Tobita, and J. Iwasa. 2000. Repair of articular cartilage defects with cultured chondrocytes in Atelocollagen gel. Comparison with cultured chondrocytes in suspension. Arch Orthop Trauma Surg. 120:121-7. Kaufmann, K., and G. Thiel. 2002. Epidermal growth factor and thrombin induced proliferation of immortalized human keratinocytes is coupled to the synthesis of Egr-1, a zinc finger transcriptional regulator. J Cell Biochem. 85:381-91.

Page 184

Kawakami, Y., J. Rodriguez-Leon, and J.C. Belmonte. 2006. The role of TGFbetas and Sox9 during limb chondrogenesis. Curr Opin Cell Biol. 18:723-9. Kawasaki, K., M. Ochi, Y. Uchio, N. Adachi, and M. Matsusaki. 1999. Hyaluronic acid enhances proliferation and chondroitin sulfate synthesis in cultured chondrocytes embedded in collagen gels. J Cell Physiol. 179:142-8. Kim, G., M. Okumura, D. Bosnakovski, T. Ishiguro, T. Kadosawa, and T. Fujinaga. 2003. Biological properties of allogenic articular chondrocytes on the surface of bovine cartilage explants in vitro. J Vet Med A Physiol Pathol Clin Med. 50:418- 23. King, J.S., L.M. Green, B.M. Bianski, M.M. Pink, and C.M. Jobe. 2005. Shaver, bipolar radiofrequency, and saline jet instruments for cutting meniscal tissue: a comparative experimental study on sheep menisci. Arthroscopy. 21:844-50. King, P.J., T. Bryant, and T. Minas. 2002. Autologous chondrocyte implantation for chondral defects of the knee: indications and technique. J Knee Surg. 15:177-84. Kirilak, Y., N.J. Pavlos, C.R. Willers, R. Han, H. Feng, J. Xu, N. Asokananthan, G.A. Stewart, P. Henry, D. Wood, and M.H. Zheng. 2006. Fibrin sealant promotes migration and proliferation of human articular chondrocytes: Possible involvement of thrombin and protease-activated receptors. Int J Mol Med. 17:551-8. Knutsen, G., L. Engebretsen, T.C. Ludvigsen, J.O. Drogset, T. Grontvedt, E. Solheim, T. Strand, S. Roberts, V. Isaksen, and O. Johansen. 2004. Autologous chondrocyte implantation compared with microfracture in the knee. A randomized trial. J Bone Joint Surg Am. 86-A:455-64. Kolettas, E., H.I. Muir, J.C. Barrett, and T.E. Hardingham. 2001. Chondrocyte phenotype and cell survival are regulated by culture conditions and by specific cytokines through the expression of Sox-9 transcription factor. Rheumatology (Oxford). 40:1146-56. Kordas, G., J.S. Szabo, and L. Hangody. 2005. The effect of drill-hole length on the primary stability of osteochondral grafts in mosaicplasty. Orthopedics. 28:401-4. Kordas, G., J.S. Szabo, and L. Hangody. 2006. Primary stability of osteochondral grafts used in mosaicplasty. Arthroscopy. 22:414-21. Kreuz, P.C., C. Erggelet, M.R. Steinwachs, S.J. Krause, A. Lahm, P. Niemeyer, N. Ghanem, M. Uhl, and N. Sudkamp. 2006a. Is microfracture of chondral defects in the knee associated with different results in patients aged 40 years or younger? Arthroscopy. 22:1180-6.

Page 185

Kreuz, P.C., M. Steinwachs, C. Erggelet, A. Lahm, S. Krause, C. Ossendorf, D. Maier, N. Ghanem, and M. Uhl. 2007. Importance of Sports in Cartilage Regeneration After Autologous Chondrocyte Implantation: A Prospective Study With a 3- Year Follow-up. Am J Sports Med. Kreuz, P.C., M.R. Steinwachs, C. Erggelet, S.J. Krause, G. Konrad, M. Uhl, and N. Sudkamp. 2006b. Results after microfracture of full-thickness chondral defects in different compartments in the knee. Osteoarthritis Cartilage. 14:1119-25. Krishnan, S.P., J.A. Skinner, R.W. Carrington, A.M. Flanagan, T.W. Briggs, and G. Bentley. 2006. Collagen-covered autologous chondrocyte implantation for osteochondritis dissecans of the knee: TWO- TO SEVEN-YEAR RESULTS. J Bone Joint Surg Br. 88:203-5. Kuo, A.C., J.J. Rodrigo, A.H. Reddi, S. Curtiss, E. Grotkopp, and M. Chiu. 2006. Microfracture and bone morphogenetic protein 7 (BMP-7) synergistically stimulate articular cartilage repair. Osteoarthritis Cartilage. 14:1126-35. Kuroda, R., A. Usas, S. Kubo, K. Corsi, H. Peng, T. Rose, J. Cummins, F.H. Fu, and J. Huard. 2006. Cartilage repair using bone morphogenetic protein 4 and muscle- derived stem cells. Arthritis Rheum. 54:433-42. laboratories., A.c.o.p.s.f.c.p.f. 2002. ATS statement: guidelines for the six minute walk test. Am J Respir Crit Care Med. 166:111-117. LaPrade, R.F., L.S. Bursch, E.J. Olson, V. Havlas, and C.S. Carlson. 2008. Histologic and immunohistochemical characteristics of failed articular cartilage resurfacing procedures for osteochondritis of the knee: a case series. Am J Sports Med. 36:360-8. Lee, H.J., J.S. Lee, T. Chansakul, C. Yu, J.H. Elisseeff, and S.M. Yu. 2006. Collagen mimetic peptide-conjugated photopolymerizable PEG hydrogel. Biomaterials. 27:5268-76. Lee, J., E. Lee, H.Y. Kim, and Y. Son. 2007. Comparison of articular cartilage with costal cartilage in initial cell yields, degree of dedifferentiation during expansion, and their redifferentiation capacity. Biotechnol Appl Biochem. Lindahl, A., M. Brittberg, and L. Peterson. 2001. Health economics benefits following autologous chondrocyte transplantation for patients with focal chondral lesions of the knee. Knee Surg Sports Traumatol Arthrosc. 9:358-63. Lisignoli, G., S. Cristino, A. Piacentini, C. Cavallo, A.I. Caplan, and A. Facchini. 2006. Hyaluronan-based polymer scaffold modulates the expression of inflammatory

Page 186

and degradative factors in mesenchymal stem cells: Involvement of Cd44 and Cd54. J Cell Physiol. 207:364-73. Louisia, S., P. Beaufils, M. Katabi, and H. Robert. 2003. Transchondral drilling for osteochondritis dissecans of the medial condyle of the knee. Knee Surg Sports Traumatol Arthrosc. 11:33-9. Lu, Y., H.D. Adkisson, J. Bogdanske, V. Kalscheur, W. Maloney, R. Cheung, A.J. Grodzinsky, K.A. Hruska, and M.D. Markel. 2005. In vivo transplantation of neonatal ovine neocartilage allografts: determining the effectiveness of tissue transglutaminase. J Knee Surg. 18:31-42. Mackie, E.J., C.N. Pagel, R. Smith, M.R. de Niese, S.J. Song, and R.N. Pike. 2002. Protease-activated receptors: a means of converting extracellular proteolysis into intracellular signals. IUBMB Life. 53:277-81. Mainil-Varlet, P., T. Aigner, M. Brittberg, P. Bullough, A. Hollander, E. Hunziker, R. Kandel, S. Nehrer, K. Pritzker, S. Roberts, and E. Stauffer. 2003. Histological assessment of cartilage repair: a report by the Histology Endpoint Committee of the International Cartilage Repair Society (ICRS). J Bone Joint Surg Am. 85-A Suppl 2:45-57. Malinin, T., and E.A. Ouellette. 2000. Articular cartilage nutrition is mediated by subchondral bone: a long-term autograft study in baboons. Osteoarthritis Cartilage. 8:483-91. Manolagas, S.C., and R.L. Jilka. 1995. Bone marrow, cytokines, and bone remodeling. Emerging insights into the pathophysiology of osteoporosis. N Engl J Med. 332:305-11. Marcacci, M., M. Berruto, D. Brocchetta, A. Delcogliano, D. Ghinelli, A. Gobbi, E. Kon, L. Pederzini, D. Rosa, G.L. Sacchetti, G. Stefani, and S. Zanasi. 2005. Articular cartilage engineering with Hyalograft C: 3-year clinical results. Clin Orthop Relat Res:96-105. Marcacci, M., S. Zaffagnini, E. Kon, A. Visani, F. Iacono, and I. Loreti. 2002. Arthroscopic autologous chondrocyte transplantation: technical note. Knee Surg Sports Traumatol Arthrosc. 10:154-9. March, L.M., and C.J. Bachmeier. 1997. Economics of osteoarthritis: a global perspective. Baillieres Clin Rheumatol. 11:817-34. Marlovits, S., P. Singer, P. Zeller, I. Mandl, J. Haller, and S. Trattnig. 2006. Magnetic resonance observation of cartilage repair tissue (MOCART) for the evaluation of autologous chondrocyte transplantation: determination of interobserver

Page 187

variability and correlation to clinical outcome after 2 years. Eur J Radiol. 57:16- 23. Marlovits, S., G. Striessnig, F. Kutscha-Lissberg, C. Resinger, S.M. Aldrian, V. Vecsei, and S. Trattnig. 2005. Early postoperative adherence of matrix-induced autologous chondrocyte implantation for the treatment of full-thickness cartilage defects of the femoral condyle. Knee Surg Sports Traumatol Arthrosc. 13:451-7. Marlovits, S., G. Striessnig, C.T. Resinger, S.M. Aldrian, V. Vecsei, H. Imhof, and S. Trattnig. 2004. Definition of pertinent parameters for the evaluation of articular cartilage repair tissue with high-resolution magnetic resonance imaging. Eur J Radiol. 52:310-9. Marlovits, S., B. Tichy, M. Truppe, D. Gruber, and W. Schlegel. 2003. Collagen expression in tissue engineered cartilage of aged human articular chondrocytes in a rotating bioreactor. Int J Artif Organs. 26:319-30. Masuoka, K., T. Asazuma, M. Ishihara, M. Sato, H. Hattori, Y. Yoshihara, T. Matsui, B. Takase, M. Kikuchi, and K. Nemoto. 2005. Tissue engineering of articular cartilage using an allograft of cultured chondrocytes in a membrane-sealed atelocollagen honeycomb-shaped scaffold (ACHMS scaffold). J Biomed Mater Res B Appl Biomater. 75:177-84. Matsusue, Y., T. Yamamuro, and H. Hama. 1993. Arthroscopic multiple osteochondral transplantation to the chondral defect in the knee associated with anterior cruciate ligament disruption. Arthroscopy. 9:318-21. Mauck, R.L., C.C. Wang, E.S. Oswald, G.A. Ateshian, and C.T. Hung. 2003. The role of cell seeding density and nutrient supply for articular cartilage tissue engineering with deformational loading. Osteoarthritis Cartilage. 11:879-90. McGraw, K., and S. Wong. 1996. Forming inferences about some intraclass correlation coefficients. Psychol Method. 1:30–46. Meister, K., A. Cobb, and G. Bentley. 1998. Treatment of painful articular cartilage defects of the patella by carbon-fibre implants. J Bone Joint Surg Br. 80:965-70. Miao, Z., J. Jin, L. Chen, J. Zhu, W. Huang, J. Zhao, H. Qian, and X. Zhang. 2006. Isolation of mesenchymal stem cells from human placenta: comparison with human bone marrow mesenchymal stem cells. Cell Biol Int. 30:681-7. Micheli, L.J., J.B. Moseley, A.F. Anderson, J.E. Browne, C. Erggelet, R. Arciero, F.H. Fu, and B.R. Mandelbaum. 2006. Articular cartilage defects of the distal femur in children and adolescents: treatment with autologous chondrocyte implantation. J Pediatr Orthop. 26:455-60.

Page 188

Miller, K. 2005. Method of testing very soft biological tissues in compression. J Biomech. 38:153-158. Minas, T. 2001. Autologous chondrocyte implantation for focal chondral defects of the knee. Clin Orthop Relat Res:S349-61. Minas, T., and T. Bryant. 2005. The role of autologous chondrocyte implantation in the patellofemoral joint. Clin Orthop Relat Res:30-9. Minas, T., and S. Nehrer. 1997. Current concepts in the treatment of articular cartilage defects. Orthopedics. 20:525-38. Minns, R.J., and M. Flynn. 1978. Intra-articular implant of filamentous carbon fibre in the experimental animal. J Bioeng. 2:279-86. Minns, R.J., and D.S. Muckle. 1989. Mechanical and histological response of carbon fibre pads implanted in the rabbit patella. Biomaterials. 10:273-6. Molino, M., E.S. Barnathan, R. Numerof, J. Clark, M. Dreyer, A. Cumashi, J.A. Hoxie, N. Schechter, M. Woolkalis, and L.F. Brass. 1997. Interactions of mast cell tryptase with thrombin receptors and PAR-2. J Biol Chem. 272:4043-9. Morel, V., and T.M. Quinn. 2004. Short-term changes in cell and matrix damage following mechanical injury of articular cartilage explants and modelling of microphysical mediators. Biorheology. 41:509-19. Moriya, T., Y. Wada, A. Watanabe, T. Sasho, K. Nakagawa, P. Mainil-Varlet, and H. Moriya. 2007. Evaluation of reparative cartilage after autologous chondrocyte implantation for osteochondritis dissecans: histology, biochemistry, and MR imaging. J Orthop Sci. 12:265-73. Moseley, J.B., K. O'Malley, N.J. Petersen, T.J. Menke, B.A. Brody, D.H. Kuykendall, J.C. Hollingsworth, C.M. Ashton, and N.P. Wray. 2002. A controlled trial of arthroscopic surgery for osteoarthritis of the knee. N Engl J Med. 347:81-8. Nathan, S., S. Das De, A. Thambyah, C. Fen, J. Goh, and E.H. Lee. 2003. Cell-based therapy in the repair of osteochondral defects: a novel use for adipose tissue. Tissue Eng. 9:733-44. Nehls, V., and R. Herrmann. 1996. The configuration of fibrin clots determines capillary morphogenesis and endothelial cell migration. Microvasc Res. 51:347- 64. Nehrer, S., H.A. Breinan, A. Ramappa, H.P. Hsu, T. Minas, S. Shortkroff, C.B. Sledge, I.V. Yannas, and M. Spector. 1998. Chondrocyte-seeded collagen matrices implanted in a chondral defect in a canine model. Biomaterials. 19:2313-28.

Page 189

Nehrer, S., S. Domayer, R. Dorotka, K. Schatz, U. Bindreiter, and R. Kotz. 2006. Three- year clinical outcome after chondrocyte transplantation using a hyaluronan matrix for cartilage repair. Eur J Radiol. 57:3-8. Nehrer, S., M. Spector, and T. Minas. 1999. Histologic analysis of tissue after failed cartilage repair procedures. Clin Orthop Relat Res:149-62. Neovius, E.B., and G. Kratz. 2003. Tissue engineering by cocultivating human elastic chondrocytes and keratinocytes. Tissue Eng. 9:365-9. Newman, A.P. 1998. Articular cartilage repair. Am J Sports Med. 26:309-24. Nomura, E., M. Inoue, and M. Kurimura. 2003. Chondral and osteochondral injuries associated with acute patellar dislocation. Arthroscopy. 19:717-21. Noth, U., R. Tuli, A.M. Osyczka, K.G. Danielson, and R.S. Tuan. 2002. In vitro engineered cartilage constructs produced by press-coating biodegradable polymer with human mesenchymal stem cells. Tissue Eng. 8:131-44. Noyes, F.R., P.A. Mooar, D.S. Matthews, and D.L. Butler. 1983. The symptomatic anterior cruciate-deficient knee. Part I: the long-term functional disability in athletically active individuals. J Bone Joint Surg Am. 65:154-62. Obedian, R.S., and R.P. Grelsamer. 1997. Osteochondritis dissecans of the distal femur and patella. Clin Sports Med. 16:157-74. O'Brien, P.J., M. Molino, M. Kahn, and L.F. Brass. 2001. Protease activated receptors: theme and variations. Oncogene. 20:1570-81. Ochi, K., A. Derfoul, and R.S. Tuan. 2006. A predominantly articular cartilage- associated gene, SCRG1, is induced by glucocorticoid and stimulates chondrogenesis in vitro. Osteoarthritis Cartilage. 14:30-8. Ochi, M., Y. Uchio, K. Kawasaki, S. Wakitani, and J. Iwasa. 2002. Transplantation of cartilage-like tissue made by tissue engineering in the treatment of cartilage defects of the knee. J Bone Joint Surg Br. 84:571-8. Ono, K., J. Shikata, K. Shimizu, and T. Yamamuro. 1992. Bone-fibrin mixture in spinal surgery. Clin Orthop Relat Res:133-9. Oreffo, R.O., S. Bord, and J.T. Triffitt. 1998. Skeletal progenitor cells and ageing human populations. Clin Sci (Lond). 94:549-55. Outerbridge, R.E. 1961. The etiology of chondromalacia patellae. J Bone Joint Surg Br. 43-B:752-7. Owen, W.G. 2003. PAR-3 is a low-affinity substrate, high affinity effector of thrombin. Biochem Biophys Res Commun. 305:166-8.

Page 190

Pagel, C.N., M.R. de Niese, L.A. Abraham, C. Chinni, S.J. Song, R.N. Pike, and E.J. Mackie. 2003. Inhibition of osteoblast apoptosis by thrombin. Bone. 33:733-43. Pavesio, A., G. Abatangelo, A. Borrione, D. Brocchetta, A.P. Hollander, E. Kon, F. Torasso, S. Zanasi, and M. Marcacci. 2003. Hyaluronan-based scaffolds (Hyalograft C) in the treatment of knee cartilage defects: preliminary clinical findings. Novartis Found Symp. 249:203-17; discussion 229-33, 234-8, 239-41. Pavlos, N.J., J. Xu, D. Riedel, J.S. Yeoh, S.L. Teitelbaum, J.M. Papadimitriou, R. Jahn, F.P. Ross, and M.H. Zheng. 2005. Rab3D Regulates a Novel Vesicular Trafficking Pathway That Is Required for Osteoclastic Bone Resorption. Mol Cell Biol. 25:5253-69. Peterson, L., M. Brittberg, I. Kiviranta, E.L. Akerlund, and A. Lindahl. 2002. Autologous chondrocyte transplantation. Biomechanics and long-term durability. Am J Sports Med. 30:2-12. Peterson, L., D. Menche, and D. Grande. 1984. Chondrocyte transplantation - an experimental rabbit in the rabbit. In 30th Annual Orthopaedic Research Society. Orthopaedic Research Society, Atlanta. 218. Peterson, L., T. Minas, M. Brittberg, A. Nilsson, E. Sjogren-Jansson, and A. Lindahl. 2000. Two- to 9-year outcome after autologous chondrocyte transplantation of the knee. Clin Orthop Relat Res:212-34. Polster, J., and M. Recht. 2005. Postoperative MR evaluation of chondral repair in the knee. Eur J Radiol. 54:206-13. Poole, A.R. 2003. What type of cartilage repair are we attempting to attain? J Bone Joint Surg Am. 85-A Suppl 2:40-4. Powers, C.M., Y.J. Chen, I. Scher, and T.Q. Lee. 2006. The influence of patellofemoral joint contact geometry on the modeling of three dimensional patellofemoral joint forces. J Biomech. 39:2783-91. Powers, J.C., and C.M. Kam. 1992. Synthetic substrates and inhibitors of thrombin. Plenum Press, New York. 117-58 pp. Prakash, D., and D. Learmonth. 2002. Natural progression of osteo-chondral defect in the femoral condyle. Knee. 9:7-10. Qiu, Y.S., B.F. Shahgaldi, W.J. Revell, and F.W. Heatley. 2002. Evaluation of Gateshead carbon fibre rod as an implant material for repair of osteochondral defects: a morphological and mechanical study in the rabbit knee. Biomaterials. 23:3943-55.

Page 191

Quinn, T.M., A.J. Grodzinsky, E.B. Hunziker, and J.D. Sandy. 1998. Effects of injurious compression on matrix turnover around individual cells in calf articular cartilage explants. J Orthop Res. 16:490-9. Recht, M., V. Bobic, D. Burstein, D. Disler, G. Gold, M. Gray, J. Kramer, P. Lang, T. McCauley, and C. Winalski. 2001. Magnetic resonance imaging of articular cartilage. Clin Orthop Relat Res:S379-96. Richardson, J.B., B. Caterson, E.H. Evans, B.A. Ashton, and S. Roberts. 1999. Repair of human articular cartilage after implantation of autologous chondrocytes. J Bone Joint Surg Br. 81:1064-8. Roberts, S., A.P. Hollander, B. Caterson, J. Menage, and J.B. Richardson. 2001. Matrix turnover in human cartilage repair tissue in autologous chondrocyte implantation. Arthritis Rheum. 44:2586-98. Roberts, S., I.W. McCall, A.J. Darby, J. Menage, H. Evans, P.E. Harrison, and J.B. Richardson. 2003. Autologous chondrocyte implantation for cartilage repair: monitoring its success by magnetic resonance imaging and histology. Arthritis Res Ther. 5:R60-73. Robertson, W.B., D. Fick, D.J. Wood, J.M. Linklater, M.H. Zheng, and T.R. Ackland. 2007. MRI and clinical evaluation of collagen-covered autologous chondrocyte implantation (CACI) at two years. Knee. 14:117-27. Robertson WB, G.H., Ackland T. 2004. Standard Practice Exercise Rehabilitation Protocols for Matrix Induced Autologous Chondrocyte Implantation Femoral Condyles. In Hollywood Functional Rehabilitation Clinic Publications, Perth, Western Australia. Robinson, D., M. Efrat, D.G. Mendes, N. Halperin, and Z. Nevo. 1993. Implants composed of carbon fiber mesh and bone-marrow-derived, chondrocyte- enriched cultures for joint surface reconstruction. Bull Hosp Jt Dis. 53:75-82. Robinson, D., and Z. Nevo. 2001. Articular Cartilage Chondrocytes are more Advantageous for Generating Hyaline-like Cartilage than Mesenchymal Cells Isolated from Microfracture Repairs. Cell Tissue Bank. 2:23-30. Ronga, M., F.A. Grassi, and P. Bulgheroni. 2004. Arthroscopic autologous chondrocyte implantation for the treatment of a chondral defect in the tibial plateau of the knee. Arthroscopy. 20:79-84. Roos, E.M., and L.S. Lohmander. 2003. The Knee injury and Osteoarthritis Outcome Score (KOOS): from joint injury to osteoarthritis. Health Qual Life Outcomes. 1:64.

Page 192

Roos, E.M., H.P. Roos, L.S. Lohmander, C. Ekdahl, and B.D. Beynnon. 1998. Knee Injury and Osteoarthritis Outcome Score (KOOS)--development of a self- administered outcome measure. J Orthop Sports Phys Ther. 28:88-96. Ruano-Ravina, A., and M. Jato Diaz. 2006. Autologous chondrocyte implantation: a systematic review. Osteoarthritis Cartilage. 14:47-51. Ruchlin, H.S., E.B. Elkin, and S.A. Paget. 1997. Assessing cost-effectiveness analyses in rheumatoid arthritis and osteoarthritis. Arthritis Care Res. 10:413-21. Rudert, M., U. Wilms, M. Hoberg, and C.J. Wirth. 2005. Cell-based treatment of osteochondral defects in the rabbit knee with natural and synthetic matrices: cellular seeding determines the outcome. Arch Orthop Trauma Surg. 125:598- 608. Russlies, M., P. Behrens, L. Wunsch, J. Gille, and E.M. Ehlers. 2002. A cell-seeded biocomposite for cartilage repair. Ann Anat. 184:317-23. Sah, R.L., Y.L. Kim, A.J. Grodzinsky, A.H.K. Plaas, and J.D. Sandy. 1992. Effects of static and dynamic compression on cartilage metabolism in cartilage explants. Raven Press, New York. 373-92 pp. Saini, S., and T.M. Wick. 2003. Concentric cylinder bioreactor for production of tissue engineered cartilage: effect of seeding density and hydrodynamic loading on construct development. Biotechnol Prog. 19:510-21. Saris, D.B., J. Vanlauwe, J. Victor, M. Haspl, M. Bohnsack, Y. Fortems, B. Vandekerckhove, K.F. Almqvist, T. Claes, F. Handelberg, K. Lagae, J. van der Bauwhede, H. Vandenneucker, K.G. Yang, M. Jelic, R. Verdonk, N. Veulemans, J. Bellemans, and F.P. Luyten. 2008. Characterized chondrocyte implantation results in better structural repair when treating symptomatic cartilage defects of the knee in a randomized controlled trial versus microfracture. Am J Sports Med. 36:235-46. Schimmer, R.C., K.B. Brulhart, C. Duff, and W. Glinz. 1998. Arthroscopic partial meniscectomy: a 12-year follow-up and two-step evaluation of the long-term course. Arthroscopy. 14:136-42. Schmitt, B., J. Ringe, T. Haupl, M. Notter, R. Manz, G.R. Burmester, M. Sittinger, and C. Kaps. 2003. BMP2 initiates chondrogenic lineage development of adult human mesenchymal stem cells in high-density culture. Differentiation. 71:567- 77.

Page 193

Schneider, U., and S. Anderereya. 2003. First results of a prospective randomised trial of traditional autologous chondrocyte transplantation and CaReS technology. Zeitschrift fur Orthopadie und Ihre Grenzgeleit. 141. Sekiya, I., K. Tsuji, P. Koopman, H. Watanabe, Y. Yamada, K. Shinomiya, A. Nifuji, and M. Noda. 2000. SOX9 enhances aggrecan gene promoter/enhancer activity and is up-regulated by retinoic acid in a cartilage-derived cell line, TC6. J Biol Chem. 275:10738-44. Shah, M.A., A.M. Ebert, and W.E. Sanders. 2002. Fibrin glue fixation of a digital osteochondral fracture: case report and review of the literature. J Hand Surg [Am]. 27:464-9. Sharma, B., C.G. Williams, M. Khan, P. Manson, and J.H. Elisseeff. 2007. In vivo chondrogenesis of mesenchymal stem cells in a photopolymerized hydrogel. Plast Reconstr Surg. 119:112-20. Smith, A. 1965. Survival of frozen chondrocytes isolated from the cartilage of adult mammals. Nature. 205:782. Sohn, D.H., L.M. Lottman, L.Y. Lum, S.G. Kim, R.A. Pedowitz, R.D. Coutts, and R.L. Sah. 2002. Effect of gravity on localization of chondrocytes implanted in cartilage defects. Clin Orthop Relat Res:254-62. Song, S.J., C.N. Pagel, T.M. Campbell, R.N. Pike, and E.J. Mackie. 2005. The role of protease-activated receptor-1 in bone healing. Am J Pathol. 166:857-68. Spahn, G., R. Frober, and W. Linss. 2006. Treatment of chondral defects by hydro jet. Results of a preliminary scanning electron microscopic evaluation. Arch Orthop Trauma Surg. 126:223-7. Star, M.J., K.R. Kaufman, S.E. Irby, and C.W. Colwell, Jr. 1996. The effects of patellar thickness on patellofemoral forces after resurfacing. Clin Orthop Relat Res:279- 84. Steadman, J.R., K.K. Briggs, J.J. Rodrigo, M.S. Kocher, T.J. Gill, and W.G. Rodkey. 2003. Outcomes of microfracture for traumatic chondral defects of the knee: average 11-year follow-up. Arthroscopy. 19:477-84. Steadman, J.R., W.G. Rodkey, S.B. Singleton, and K.K. Briggs. 1997. Microfracture technique for fullthickness chondral defects: technique and clinical results. Operat Tech Orthop. 7:300-304. Takahashi, T., B. Tins, I.W. McCall, J.B. Richardson, K. Takagi, and K. Ashton. 2006. MR appearance of autologous chondrocyte implantation in the knee: correlation with the knee features and clinical outcome. Skeletal Radiol. 35:16-26.

Page 194

Tanaka, N., T. Morita, A. Nezu, A. Tanimura, I. Mizoguchi, and Y. Tojyo. 2003. Thrombin-induced Ca2+ mobilization in human gingival fibroblasts is mediated by protease-activated receptor-1 (PAR-1). Life Sci. 73:301-10. Tani, K., S. Yasuoka, F. Ogushi, K. Asada, K. Fujisawa, T. Ozaki, N. Sano, and T. Ogura. 1991. Thrombin enhances lung fibroblast proliferation in bleomycin- induced pulmonary fibrosis. Am J Respir Cell Mol Biol. 5:34-40. Tatebe, M., R. Nakamura, H. Kagami, K. Okada, and M. Ueda. 2005. Differentiation of transplanted mesenchymal stem cells in a large osteochondral defect in rabbit. Cytotherapy. 7:520-30. Tegner, Y., and J. Lysholm. 1985. Rating systems in the evaluation of knee ligament injuries. Clin Orthop Relat Res:43-9. Teixeira, J.O., and M.R. Urist. 1998. Bone morphogenetic protein induced repair of compartmentalized segmental diaphyseal defects. Arch Orthop Trauma Surg. 117:27-34. Tew, S.R., A.P. Kwan, A. Hann, B.M. Thomson, and C.W. Archer. 2000. The reactions of articular cartilage to experimental wounding: role of apoptosis. Arthritis Rheum. 43:215-25. Tigenix. 2007. TiGenix announces positive phase III trial results. In Press Release. G. Beyen and K. Motmans, editors. http://www.tigenix.com, Belgium. Tins, B.J., I.W. McCall, T. Takahashi, V. Cassar-Pullicino, S. Roberts, B. Ashton, and J. Richardson. 2005. Autologous chondrocyte implantation in knee joint: MR imaging and histologic features at 1-year follow-up. Radiology. 234:501-8. Tippet, J. 1991. Articular cartilage drilling and osteotomy in osteoarthritis of the knee. Raven Press, New York. 325-39 pp. Torzilli, P.A., R. Grigiene, C. Huang, S.M. Friedman, S.B. Doty, A.L. Boskey, and G. Lust. 1997. Characterization of cartilage metabolic response to static and dynamic stress using a mechanical explant test system. J Biomech. 30:1-9. Trattnig, S., A. Ba-Ssalamah, K. Pinker, C. Plank, V. Vecsei, and S. Marlovits. 2005. Matrix-based autologous chondrocyte implantation for cartilage repair: noninvasive monitoring by high-resolution magnetic resonance imaging. Magn Reson Imaging. 23:779-87. Ueblacker, P., A. Burkart, and A.B. Imhoff. 2004. Retrograde cartilage transplantation on the proximal and distal tibia. Arthroscopy. 20:73-8.

Page 195

Ueno, T., T. Kagawa, N. Mizukawa, H. Nakamura, T. Sugahara, and T. Yamamoto. 2001. Cellular origin of endochondral ossification from grafted periosteum. Anat Rec. 264:348-57. Visna, P., L. Pasa, J. Adler, J. Folvarsky, and D. Horky. 2003a. [Treatment of deep chondral defects of the knee using autologous chondrocytes cultured on a support--preparation of the cartilage graft]. Acta Chir Orthop Traumatol Cech. 70:350-5. Visna, P., L. Pasa, R. Hart, J. Kocis, I. Cizmar, and J. Adler. 2003b. [Treatment of deep chondral defects of the knee using autologous chondrocytes cultured on a support--results after one year]. Acta Chir Orthop Traumatol Cech. 70:356-62. Vu, T.K., D.T. Hung, V.I. Wheaton, and S.R. Coughlin. 1991. Molecular cloning of a functional thrombin receptor reveals a novel proteolytic mechanism of receptor activation. Cell. 64:1057-68. Wakitani, S., T. Goto, S.J. Pineda, R.G. Young, J.M. Mansour, A.I. Caplan, and V.M. Goldberg. 1994. Mesenchymal cell-based repair of large, full-thickness defects of articular cartilage. J Bone Joint Surg Am. 76:579-92. Wakitani, S., T. Kimura, A. Hirooka, T. Ochi, M. Yoneda, N. Yasui, H. Owaki, and K. Ono. 1989. Repair of rabbit articular surfaces with allograft chondrocytes embedded in collagen gel. J Bone Joint Surg Br. 71:74-80. Waldman, S.D., C.G. Spiteri, M.D. Grynpas, R.M. Pilliar, J. Hong, and R.A. Kandel. 2003. Effect of biomechanical conditioning on cartilaginous tissue formation in vitro. J Bone Joint Surg Am. 85-A Suppl 2:101-5. Wang, H.S., F. Li, M.S. Runge, and E.L. Chaikof. 1997. Endothelial cells exhibit differential chemokinetic and mitogenic responsiveness to alpha-thrombin. J Surg Res. 68:139-44. Wasiak, J., C. Clar, and E. Villanueva. 2006. Autologous cartilage implantation for full thickness articular cartilage defects of the knee. Cochrane Database Syst Rev. 3:CD003323. Weinand, C., G.M. Peretti, S.B. Adams, Jr., L.J. Bonassar, M.A. Randolph, and T.J. Gill. 2006. An allogenic cell-based implant for meniscal lesions. Am J Sports Med. 34:1779-89. Whiteside, R.A., R.P. Jakob, U.P. Wyss, and P. Mainil-Varlet. 2005. Impact loading of articular cartilage during transplantation of osteochondral autograft. J Bone Joint Surg Br. 87:1285-91.

Page 196

Wickham, M.Q., G.R. Erickson, J.M. Gimble, T.P. Vail, and F. Guilak. 2003. Multipotent stromal cells derived from the infrapatellar fat pad of the knee. Clin Orthop Relat Res:196-212. Wildner, M., O. Sangha, and C. Behrend. 2000. Wirtschaftlichkeitsuntersuchung zur autologen Chondrozytentransplantation. Arthroskopie. 13:123-31. Willers, C., J. Chen, D. Wood, J. Xu, and M.H. Zheng. 2005. Autologous chondrocyte implantation with collagen bioscaffold for the treatment of osteochondral defects in rabbits. Tissue Eng. 11:1065-76. Willers, C., T. Partsalis, and M.H. Zheng. 2007. Articular cartilage repair: procedures versus products. Expert Rev Med Devices. 4:373-392. Willers, C., D. Wood, and M. Zheng. 2003. A current review on the biology and treatment of articular carttilage defects (part I & part II). J Musculoskeletal Res. 7:157-181. Wood, J.J., M.A. Malek, F.J. Frassica, J.A. Polder, A.K. Mohan, E.T. Bloom, M.M. Braun, and T.R. Cote. 2006. Autologous cultured chondrocytes: adverse events reported to the United States Food and Drug Administration. J Bone Joint Surg Am. 88:503-7. Wu, J.Z., W. Herzog, and E.M. Hasler. 2002. Inadequate placement of osteochondral plugs may induce abnormal stress-strain distributions in articular cartilage -- finite element simulations. Med Eng Phys. 24:85-97. Yao, X., X. Ma, and Z. Zhang. 2000. [The repairment of the condylar cartilage defect by transplantation of chondrocytes embedded in the collagen membrane]. Zhonghua Kou Qiang Yi Xue Za Zhi. 35:138-41. Yates, J.W., Jr. 2003. The effectiveness of autologous chondrocyte implantation for treatment of full-thickness articular cartilage lesions in workers' compensation patients. Orthopedics. 26:295-300; discussion 300-1. Yelin, E. 2003. Cost of musculoskeletal diseases: impact of work disability and functional decline. J Rheumatol Suppl. 68:8-11. Yokoyama, A., I. Sekiya, K. Miyazaki, S. Ichinose, Y. Hata, and T. Muneta. 2005. In vitro cartilage formation of composites of synovium-derived mesenchymal stem cells with collagen gel. Cell Tissue Res. 322:289-98. Zheng, M., C. Willers, L. Kirilak, P. Yates, J. Xu, D. Wood, and A. Shimmin. 2007a. Matrix-induced Autologous Chondrocyte Implantation (MACI®): Biological and Histological Assessment. Tissue Engineering. In Press.

Page 197

Zheng, M.H., E. King, Y. Kirilak, L. Huang, J.M. Papadimitriou, D.J. Wood, and J. Xu. 2004. Molecular characterisation of chondrocytes in autologous chondrocyte implantation. Int J Mol Med. 13:623-8. Zheng, M.H., R. Pembrey, S. Niutta, P. Stewart-Richardson, and A. Farrugia. 2006. Challenges in the evaluation of safety and efficacy of human tissue and cell based products. ANZ J Surg. 76:843-9. Zheng, M.H., C. Willers, L. Kirilak, P. Yates, J. Xu, D. Wood, and A. Shimmin. 2007b. Matrix-induced autologous chondrocyte implantation (MACI): biological and histological assessment. Tissue Eng. 13:737-46.

VVÜÜtt||zz jj||ÄÄÄÄxxÜÜáá EECCCCKK