comparison of phosphate scaffolds for candidate … · bone tissue engineering 1. b .2.a...

181
A COMPARISON OF 3-DIMENSIONAL CALCIUM PHOSPHATE SCAFFOLDS FOR CANDIDATE BONE TISSUE ENGINEERING CONSTRUCTS Dolores Ba ks h A thesis submitted in conformity with the requirements for the degree of Master of Applied Science Department of Chexnical Engineering and Applied Chemistry & The Institute of Biomaterial and Biomedical Engineering University ofToronto O Copyright by Dolores Baksh 1999

Upload: others

Post on 21-Jul-2020

2 views

Category:

Documents


0 download

TRANSCRIPT

A COMPARISON OF 3-DIMENSIONAL CALCIUM PHOSPHATE SCAFFOLDS FOR CANDIDATE BONE TISSUE ENGINEERING

CONSTRUCTS

Dolores Ba ks h

A thesis submitted in conformity with the requirements for the degree of Master of Applied Science

Department of Chexnical Engineering and Applied Chemistry & The Institute of Biomaterial and Biomedical Engineering

University of Toronto

O Copyright by Dolores Baksh 1999

National Library Bibliit m u e nationale du Canada

Acquisitions and Acquisitions et Bi bliographic Services services bibliographiques 395 Wellington Street 395. rue Weîlington OttawaON KlAON4 OmwaON KlAûN4 canada Canada

The author has granted a non- exclusive licence allowing the National Library of Canada to reproduce, loan, distribute or sell copies of this thesis in microfoxm, paper or electronic formats.

The author retains ownership of the copyright in this thesis. Neither the thesis nor substantial extracts fkom it may be printed or othenvise reproduced without the author's permission.

L'auteur a accordé une licence non exclusive permettant à la Bibliothèque nationale du Canada de reproduire, prêter, distribuer ou vendre des copies de cette thèse sous la forme de microfiche/fÏltn., de reproduction sur papier ou sur format électronique.

L'auteur conserve la propriété du droit d'auteur qui protège cette thèse. Ni la thèse ni des extraits substantiels de celle-ci ne doivent être imprimés ou autrement reproduits sans son autorisation.

Dolores Baksh, M.A.Sc., 1999 Department of Chemical Engineering and Applied Chemistry & The Lnstitute for Biomaterials and Biomedical Engineering

University of Toronto

Many materids are available for osseous repair of which calcium phosphates (CP)

are generally considered materials of choice, and have been adopted as sca£folds for the

restoration of bone stock through bone t h e engineering (TE) strategies. However, the

ided CP scaffold for a bone TE constnict has yet to be found. It was hypothesized that

one essential criterion for a successfid CP-TE scafZold is a fully intercomecting

macroporosity that would permit three dimensional tissue growth. Thus, five difTerent

vpes of porous CP scaffolds, obtained fiom Zimmer international Inc., CAM Implants,

and School of Materials Engineering, Yeungnam University, Korea, ['CPP'], having

different macroporosities, were investigated for their suitabiiity as TE constructs.

The highest degree of intercomecting macroporosity was found in the CPP

scaEold-types. The Zimmer and CAM implant scaffolds possessed macropores but linle,

or no, interconnecting macroporosity. Al1 CPs supported bone growth in viîro. However,

the CPP scafZolds demonstrated bone growth throughout their porous network, while

bone growth was on the Zimxner and CAM Implant scaffolds was restricted to their outer

surfaces.

ACKNOWLEDGrnNTS

1 wish to thank Dr. John E. Davies, otherwise known as JED, for giving me the

opportunity to be a part of the infamous Bone Interface Group. With your advice, unique

scientrfic perspective and insight, my time at the Centre for Biomaterials has been most

mernorable and educational. JED's great enthusiasm to teach and help bas made tbis

work possible.

1 am grateful to Amy Shiga, for not only her technical assistance and expertise

surrounding various aspects of my work but more irnportantly, for her firiendship which 1

have come to value. 1 wish also to express my appreciation to Robert Chemecky for

teaching me SEM techniques required to produce the high quality micrographs that are

included in this work. Sincere thanks to ali members of Dr. Davies' Bone Interface

Group, Shen, Raisa, Chantal, Elaine, Soheila, Jun and Moms. Working with this diverse

group has been an invaluable experience.

To al1 my fiends at the Centre for Biomaterials, thanks for the help and

fiiendships. 1 could not imagine working in a fEendlier environment 1 would especially

like to thank Samar and Nikki who have made the day-to-day thesis grind enjoyable. Our

morning coffee at Second Cup was something 1 looked fonvard to each day - it was

worth plowing through morning rush-hour &c. Those times will be missed. We'll keep

in touch. Remember, we need to talk in 4 years!! !

To ail my &ends outside the realm of school, Carol Am, Gloria, Ildiko and

Larissa, thanks for your a m d g fiiendships throughout my studies. 1 greatly admire your

iasights and talents and wish you aII success in the fbture.

Most importantly, 1 wish to thank m y parents for their never ending understanding

and support, Your love and dedication has enabled me to strive to W l l my academic

aspirations. To my sister, Monica, you have coIlStantly reminded me of the higher power

that has given me talents that 1 should maximize. Thank you for your encouragement and

love. T o m y family, my success is your success!

D. Baksh

CONTENTS

List of Figures, Graphs & Tables

Abbreviations

1 . GENERAL INTRODUCTION

1 .A. Bone Gr& & Bone Substitutes 1

1 .A. 1. Boue Transplantation 1

1 A. 1 .a. Morphology of Bone 2

1 .A. 1. b. Function & Composition of Bone 4

1 .A. 1 .c. Bone Remodeling 6

1 .A. 1 .d. De Novo Bone Formation

1 .A.2. Bone Grafting Materials

1 .A.2.a. Autogeaic Materials 9

1 .A.2.b. Allogenic Materials

1 .A.2.c. Xenogenic Gr&

1 .A.2.d. Synthetic Bone Substitutes

1 .B - Tissue Engineering

1 .B. 1. General

1 .B.2. Bone Tissue Engineering

1. B .2.a Macrostructures of Trabecular Bone 23

1 .%.2.b. Candidate ScafToIds for Bone Tissue Engineering 26

1.B.2.c. Current Limitations of Calcium Phosphate (CP) Bone TE 29 ScaEoldS

2. RESEARCH RATiONALE

3. HYPOTHESIS

4. OBJECTIVES

5. MATERIALS & METHODS

5.A. Methods of Characterization

5.A. 1 . Provision of Calcium Phosphate Scaf3olds

5.A.2. Light Photography

5 .A.3. X-Ray Diffraction Spectroscopy @RD)

5 .A.4. Scannuig Electron Microscopy (SEM)

5.A.5. Back Scattered Electron Microscopy (BSE)

S. A.5 .a. LR White Embedding Protocois

5.AS.b. Quantification of Macroporosity and Intercomectivity

5.A.6. Degradation Behaviour of CP Scaffolds

S.B. Static In V h Studies

5.B. 1 . Rat Bone Marrow Culture System

5.B.2. Description of Static Ce11 Culture Configuration

S.B.2.a. Prùnary Ce11 Culture, Subculture & Initial Ce11 Seeding

S.B.2.b. Tissue Culture Substrate for Ceil Seeding

5.B.2.c. Ce11 Adherrence as a fimction of Seeding Time

5.B.2.d. Ce11 Adherence on 3-D Substrates during Ce11 Seeding

5.C. Dexamethasone (-) Cultures

5.D. Dynamic In Vitro Studies

5.D. 1. Description of Dynamic Ce11 Culture Configuration

S.E. Ce11 Colonization, Arrangement and Function in Static & Dynamic Culture Systems as a function of Tune

5 .F. In Vivo Studies

5 .F- 1. Surgical Procedures

5 .F.2. Histological Preparatiom

5.F.2.a. In Vivo Samples

5.F.2.b. ln Vitro Samples

6. RESULTS

6.A. Methods of Characterization

6.A. 1. Light Photography of As-received Samples

6.A.2. Powder X-Ray Diffraction @RD) Spectroscopy

6.A.3. ScannÏng Electron Microscopy

6.A.3 .a. Micro and Macroporosity

6.A.3 -b. Intercomectivity

6.A.4. Quantification of Total Porosiw

6.A.5. Degradation Behaviour of CP Scaffold~

6.AS.a. Atomic Absorption Spectroscopy

6.A.5.b. SEM of Calcium Phosphate Surfaces during Degradation Study

6.B. Static In Viîro Studies

6.8.1. Ce11 Culture Substrate for Ce11 Seeding and Colonization

6.B.2. Optimum Ce11 Seeding Time

6.B.3. Celi Adherence & Colonization on 3-D CPs during Cell Seeding

vii

6-B.4. Dexamethasone (-) Cultures

6.C. Ce11 Colonization, Arrangement and Function in Static & Dynamic Culture Systems as a fùnction of T h e

6.C. 1. SEM Observations in the Static Culture Environment

6.C .2. SEM Observations in the Dynamic Culture Environment

6.C.3. LM of Ce11 Colonization and Arrangement in Static & Dyoamic Culture Systems

6.C.4. BSE haging

6.D. In Vivo Studies

7. DISCUSSION

7.A. Physical Characterization of Candidate E ScdTolds

7.A. 1. ScafTold Composition

7.A.2. Porosity & Intercomecting Macroporosity

7.A.3. Degradation Behaviour

7.A.3 .a. Solution-mediated Processes

7.A.3 .b. Cell-mediated Processes

7.B. In Vitro Biological Characterization of Candidate TE ScaEolds

7.B.1. Ce11 Culture Techniques Involving Porous 3-D Substrates

7.B.2. Appropriate Cell Seeding T i e

7.B.3. Ce11 Colonization and Arrangement on 3-D Scaf5oids after 4 hours of Static & Dynarnic Culturing

7.B.4. Pore Bridging & Occlusion

7.B.5. Osteogenic Activity on 3-D Scaffolds maintained in Static & Dynamic Culture Systems

7.B.6. The Suitability of Dynamic Culturing

viii

7.C. In Vivo Bone Growth throughout 3-D ScafTold

8. RELATING RESULTS OF THE STUDY BACK TO THE HYPOTHESIS

9. CONCLUSIONS

Appendix A: Composition & Preparation of Fully Supplemented Medium

Appendix B: Calculation of Maximum Ce11 Seeciing Density

Appendix C: Preparation of Karnovsky's Fixative

Appendix D: Long-acting Ascorbic Acid

1 0. REFERENCES

LIST OF FIGURES, GRAPHS & TABLES

Figure 1.1

Figure 1.2

Table 1.1 Table 1.2

Figure 1.3

Figure 1.4

Table 1.3 Figure 5.1 Figure 5.2 Figures 6.1 A-B Figures 6.2A-B Figures 6.3A-B Figures 6.4A-B Figures 6.5A-B Figures 6.6A-B Figures 6.7A-B Figures 6.8A-B Table 6.1 Graphs 6.1-6.5 Figures 6.9A-P Figures 6.1 0A-1 Table 6.2 Figures 6.1 1 Figure 6.12

Table 6.3 Figure 6.13

Figure 6.15

Figure 6-16

Figure 6.1 7

Figure 6.18

Schematic view of a longitudinal section through a growing long bone Diagrammatic representation of the establishment of an interface between bone tissue and an underlying substratte Ceramic bone substitute distributed in the United States The mechanical properties of bone, metallic and calcium phosphate implant materials Light micrographs showing the isotropic and anisotropic areas of trabeculae in the neck of the femora Diagram showing the directions of the trabeculae in the femora Averages of porosity and trabeculae width Ce11 seeding on to 3-D substrates Dynamic ce11 culture apparatus Light photographs of as-received Zimmer scatfolds Light photographs of as-received CAM40/60 scaffolds Light photographs of as-received CAM70/30 scaffolds Light photographs of as-received 2CAM70130 scaffolds Light photographs of as-received CPP-6Oppi scaffolds Light photographs of as-received CPP-45ppi scafTolds Light photographs of as-received CPP-20ppi scaffolds Light photographs of as-received CPP- 1 Oppi scaffolds XRD Results of as-received Calcium Phosphate Scaffolds XRD spectra of as-received CP scaffolds SEM of as-received Zimmer and CAM Implant scaffolds SEM of as-received CPP scafTolds Micro/rnacrostructurai properties of the as-received scaffolds SEM of human trabecular bone A typical BSE image generated fiom a LRW embedded CP S cafFold Relationship between pixel intensity range and matenal type IUustration of a typical image generated in Sigmascan Pro Fom a BSE image Montage of as-received Z b e r surface afier LRW infiltration Montage of as-received CAM40160 surface after LRW infiltration Montage of as-received CAM70/30 surface d e r LRW infiltration Montage of as-received 2CAM70/30 surface &er LRW infiltration Montage of as-received CPPdOppi surface d e r LRW infiltration

Graph 6.6 Table 6.4 Graph 6.7

Table 6.5 Graph 6.8

Figures 6.1 9A-L Figures 6.20A-L Figures 6.2 1 A-1 Figures 6.22A-1

Figures 623A-1 Figures 6.24A-F

Figures 6.25A-F

Graph 6.9 Graph 6.10 Figures 6.26A-D

Figures 6.27A-F

Figures 6.28A-D

Figures 6.29A-T

Figures 6.30A-T

Figures 6.3 1 A-D Figures 6.32A-L

Figures 6.33A-D

Figures 6.34A-C

Figures 6.35A-D

Figures 6.36A-D

Figures 6.37A-D

Total porosity of as-received CP scafKolds Tabulated total porosity with corresponding rank Calcium ion leached from CP samples incubated in 0.1 M Tris buffer as a fkction of time Degradation behaviour in 0.1 M Tris bunèr (pH 7.4) Change in 0.1 M Tris buffer pH during the 6-week degradation study period SEM of Zimmer samples incubated in 0.1 M Tris buffer SEM of CAM40/6O samples incubated in O. 1 M Tris b&er SEM of CAM7OBO samples incubated in 0.1 M Tris b d e r SEM of 2CAM70/30 samples incubated in O. 1 M Tris buffer SEM of CPP-2Oppi samples incubated in 0.1 M Tris b a e r SEM of ce11 colonization and activity on 3-D porous ceramics cultured in TCT 24-well plates SEM of cell colonization and activity on 3-D porous ceramics culhrred in BG 24-well plates Total cell attachment to TCT plastic as a fiinction of tirne Total ce11 attachment to various substrates after 1 hour Colour Light photographs showing various sucfaces of the CAM40/60 and Zimmer samples sbined with toluidine blue after trypsinization SEM showing the Zimmer, CPP and CAM40/60 surfaces pst-trypshkation SEM of ce11 colonization on 3-D porous substrates d e r 1 hour of ce11 seeding SEM showing the ce11 population colonizing the surfaces of the CP samples after 1 week incubation in DEX(-) culture conditions SEM of the colonization, migration and activity on 3-D porous ceramics at 4 hours 2 days SEM of ce11 bridging at 2 days SEM of the colonization of cells on 3-D porous substrates d e r 6 weeks SEM of CPP samples cultured in DEX(+) static culture conditions at 1 week SEM of Zimmer surfaces after 1 week in DEX(+) culture conditions Cernent Iine formation in static and dynarnic culture environrnents SEM showing the extent of bone matrix elaboration on 3-D porous ceramics cultured in static and dynarnic environments SEM showing the appearance of the CPP scaffolds cultured in the presence of RBM cells for 2 weeks

Figures 6.3 8A-D

Figures 6.39A-B

Figures 6.40A-B

Figures 6.4 1 A-C

Figures 6.42A-D

Figures 6.43A-D

Figure 6.44

Figures 6.45A-D

Figures 6.46A-B

Figures 6.47A-D

Six weeks after -tic and dynamic ce11 culturing of porous CP substrates (H&E) Extent of pore bridging on 3-D porous ceramic surfaces d e r 6 weeks of static and dynamic ce11 culturing (H&E) Lack of osteogenesis within the buk of the CAM Implant samples (Azan Heidenhain comective tissue stain) Bone formation within pore volumes of CP samples cultured dynamically (H&E) Confirmation of bone formation by the Azan Heidenhain connective tissue stain Extent of bone matrix elaboration on CP scafEoIds after 8 weeks of static culturing (H&E) Montage of BSE image of the CAM40/60 surface after 6 weeks of cell cuIture SEM showing the CPP scaEolds in vivo retrieved afier 2 weeks Twenty-three weeks after transfemoral implantation of the CPP samples (H&E) Evidence of osteoclastic resorption of CPP scafToIds implanted in rat femora after 23 weeks W&E)

xii

a-MEM PGP AA BG BMD BMDC Ca CP DCC DEX (+) DEX (-) EDTA FBS FSM F.W. LM MW Od Or PU SCC SEM TB TCT TE XRD

Alpha minimal essential medium Beta glycerophosphate Ascorbic acid Bacterïological grade Bone marrow derived Bone marrow derived cells Calcium Calcium phosphate Dynamic ceU cuIture Dexamethasone supplemented culture medium Dexamethasone omitted fiom culture medium Ethylenediaminetetracetic acid Fetai bovine senun Fully supplemented medium Field width Light microscopy Molecdar weight Outer diameter Inner diameter Po lyurethane Static cell culture Scanning electron microsc Toludine blue Tissue culture treated Tissue engineering X-ray difiction

1. GENERAL INTRODUCTION

I.A. Bone Grafts & Bone Substitutes

Many investigators in the fields of medicine, dentistry and biomedical

engineering are searching for the best methods of restoring or replacing lost diseased

andor damaged bone. At present, large quantities of materials are available for osseous

repair and they can be divided into two categories: bone gr& and alloplastic implant

materials. Bone grafts include autogenous, allogenous and xenogenous grafi types-

Alloplastic impiants, in contrast to aliogenic grafts, are synthetically derived. Auto&

are preferred over the other bone grafb because they are biocompatible; that is, they do

not elicit an immune response when ùnplanted. However, the use of autografts is

primarily limited by the volume of autogenous bone tissue that is available for repair.

Allografts and xenografts do not have such limitations. However, they do cause

immunological responses and may not degrade efficiently once implanted, therefore

making them less than ideai graft matenais. Consequentiy, the more favorable types of

implant materials are alloplastics. Large selections of alloplastic implant materiais are

used in clinicai applications and they can be sub-divided into four categories: metals,

ceramics, polymers and composites. In particular, synthetic calcium phosphate (CP)

ceramics have been under investigation during the last decades for their potential use as

bone replacement material mainly because calcium phosphate salts fom the major

inorganic component of bone tissue.

1 .A. 1. Bone Transplantation

Trauma resulting in large skeletal defects, referred to as criticai-sized bony

deficits, will not regenerate spontaneously, therefore, a suitable substance to accelerate

the healing and to restore form and f"unction is required (Katthagen, 1986). Procedures

involving the use of autografts and allogenic-banked bone are performed approximately

200 000 times annuaily in the United States (Lane et al., 1996) to treat such trauma

However, there is an unacceptable failure rate associated with such graft material;

autogenic (3.5% gr& hcture, 7.1% questionable graft viability and 5% infection in their

patients) and banked bone (16.5% graft hcture and 20% failure to heal normally)

(Murphey et ai., 1992)- Considering such failure rates as well as issues associated with

potentid pathogenic transmissions and immune responses in d o g r a h , the development

of sak and efficient alternatives is curreatiy under investigation.

A bone transplant utilizing bone grafts (autograft, allograft and xenograft) is

considered successfd if the bone gr& achieves specific biologicd funçtions. First, the

gr& should facilitate osteogenesis; that is, the cells of the g r f i that survive should

produce new bone as a consequence of revascularizattion. Second, the grafi matenal

should be osteoinductive by possessing protein mediators in the matrix of the graft that

induce bone formation locaily by recruiting cells that have a potential for bone formation.

FinaIly, the gr& shouid be osteoconductive by providing a fiamework for blood vesse1

ingrowth and cells (Goldberg, 1992). The uitimate function of a bone graft is to provide

structural support. Each grafting matenal may satisQ one or several of these functions

that in turn reflect its success as a bone replacement gr&.

1.A.l.a- Morphology of Bone

There are two main types of bones that can be distinguished anatomically: flat

bones (skull bones, scapula, maqdible and ileum) and long bones (tibia, femur, and

humerus). There are two microarchitechral forms of bone: cortical (compact) and

cancelious (trabecular/spongy). Examinhg the surface external of a long bone (Figure

wt

Epip hysis

Growth Plate

Metaphysis

Cortical bone

Endosteum

Periosteum

Growth Plate

'ai ?cellous bone

Oiaphysis

Figure 1.1. Schematic view of a longitudinal section through a growing long bones (From Jee WSS. The skeletal tissues. In Weiss, L. (ed) Histology, Ce11 and Tissue Biology. Elsevier Biomedical, New York, pp 220- 255,1983)

two distinct regions are observed: two wider extremities (the epiphyses) and a cylindrical

tube in the middle (the diaphysis). There is a developmental zone that is located between

these two regions called the metaphysis. During the growth of long bones, a layer of

cartilage separates the epiphysis and the metaphysis called the growth plate. This region

contains proliferative cells and expanding cartilage rnatrix that is responsible for the

longitudinal growth of bone and at the end of the growth penod, this layer becomes

calcified and remodeled. A thick, dense layer of calcified tissue, the cortex (compact or

cortical bone), comprises the extenial part of bone. The meddary cavity, containing

hematogoietic marrow, located in the diaphysis portion of long is enclosed by compact

bone. Toward the epiphyses the compact bone becomes thinner and the intemal space

filled with a network of thin, calcified trabeculae; known as spongy or trabecular bone.

The interna1 cavity of trabecular bone is filled with hematopoietic bone marrow that is

confluent with the marrow of the medullary cavity located in the diaphysis region- During

osteoporosis, for example, a metabolic bone disease, the trabeculae become disrupted and

Iose their connectivity. The loss of co~ectivity results in skeletal fiagility (Lyndon et al.,

1996). A possible treatment of the disease is filling the osteoporotic site with a paf%

material.

There are two surfaces that bone contacts with soft tissue: the periosteal surface

and the endosteal surface. The periosteum and the endosteum are lined with osteogenic

cells organiqd in layers. The periosteum is made up of two Iayers: an outer fibrous layer

and an inner layer of soft connective tissue. The inner layer contains potential osteogenic

cells, referred to, during quiescent times, as resting or lining cells- The endosteum

comprises a layer of differentiating osteogenic cells (DOC) that are recruited for bone

synthesis.

1.A.l.b. Function & Composition of Bone

Bone is a highiy specialized f o m of connective tissue and together with cartilage,

makes up the skeletal system. Bone serves three main fùnctions: (1) mechanical support

and sites for muscle attachment for iocomotion~ (2) protection of vital organs and bone

marrow; and (3) as by providing a reserve of ions, particuiarly calcium and phosphate, for

ionic homeostasis in the body (Baron, 1996). Bone is composed of an organic matrix that

is strengthened by deposits of calcium salts and cells- Type 1 collagen comprises 95% of

the organic matrix and the remaining 5% constitutes proteoglycans and various

noncollagenous proteins. The crystalline salts incorporated into bone under the cellular

control are essentially caicium and phosphate in the form of hydroxyapatite (Marks et al.,

1996).

Bone is composed of four ciifferent types of cells: osteoblasts, osteocytes, bone

lining c e k and osteoclasts- There are two distinct lineages that bone ceils originate fiom:

osteoblasts, osteocytes and bone Iining ceils originate from local osteoprogenitor cells

and osteoclasts arise fiom the fusion of blood-borne mononuclear precursors. Osteoblasts

are fully differentiated cells that &se fiom many different stages of functional

differentiation. At each stage, the phenotype, morphological appearance and biosynthetic

activity of the differentiating osteoblast are different. Osteoblasts arise fiom pluripotent

mesenchymai stem cells of the bone marrow (Aubin et al., 1996). These stroma1 cells

have the potential to become osteoblasts as well as become fibroblasts, chondrocytes,

adipocytes or muscle ceiis. Based on morphological and histological studies (Aubin et al.,

1996), a linear sequence fiom osteoprogenitor to preosteoblasts, osteoblasts, and lining

cells or osteocytes is presumed. Osteoblasts are cuboidal plump celIs, sometimes

organized in layers, which synthesize bone matrix. An osteoblast secretes type 1 collagen

and noncollagenous proteins that comprise the organic matrix of bone. The osteocyte is a

mature osteoblast that has become embedded in bone matrix but is no longer

synthetically active. Each osteocyte is encased in a lacuna within the matrix and extends

filopodial processes through canaliculi in the matrix to make connections with adjacent

cells via gap junctions. These filopodiai connections permit communication between

adjacent osteocytes embedded in bone matrix. Canaliculi allow for the diffusion of

nutrients and metabolites. Bone lining cells are flat, elongated and inactive cells with few

cytoplasmic organelles that are situated dong bone surfaces that are not participating in

bone formation or bone resorption (Marks et al., 1996). It has k e n speculated that bone

lining cells may be precursors to osteoblasts (Marks et al., 1996). Osteoclasts are ceus

that carry out bone resorption. This bone ce11 type arises fiom the pluripotent stem cells

of the bone marrow. which generate aü blood ceus. They lïkely origuiate from the

monocyte macrophage lineage and diverge fiom the monocyte precursor (Rodan, 1992).

Osteoclasts are large, multinucleated, cells that when active rest directly on the bone

surface. In their active state, osteoclasts have two plasma membrane specialktions: a

ruffled border and a clear zone. The nifned border is the highly folded area of the plasma

membrane where resorption takes place. At the edge of the W e d border, there is a ~g

of membrane, the clear zone, which adheres tightly to the bone and seals the resorption

site.

I.A. 1.c. Bone Remodeling

Bone is a dynamic tissue that undergoes remodeling throughout life. Remodeling

is the process of forming bone in areas where bone resorption has previously occurred. At

the remodeling site, an 'activation-resorption-formation' sequence occurs where

osteoclasts, which are responsible for resorbing the bone, and osteoblasts fil1 in the

resorbed areas with bone. During osteoporosis, there is an irnbalance between the rate of

bone resorption relative to bone formation, leading to a decrease in bone mass and

structural deterioration of the skeleton (Lyndon et al., 19%).

At the interface of old bone and new bone an &brilla., noncollagenous matrix is

deposited. This matrix is referred to as the 'cernent line' or Kittlinien (Gennan) as first

described by von Ebner in 1875. Davies et al. (1991) provided morphological evidence

to suggest that osteogenic ceils are responsible for the deposition of cernent lines. They

then proceed to elaborate iuimineralized osteoid that will eventuaily be mineralized into

bone ma&. Cernent Iines can be visualized by haematoxyh staining and appear as

basophilie bands in both decalcified paraffin (Pritchard, 1972) and undecalcifieci sections

(Gruber et al., 1985). The widths of cement lines have been reported to measure between

0.2 pm to 5 prn (Villanueva et al., 1986; Philipson, 1965). Cernent lines are clearly a

bdamental occurrence for bone tissue formation that demarcates the interface of old and

new bone.

l.A.l.d. De Novo Bone Formation

There is essentiaüy a four-stage sequence of events occwrhg duruig de novo

bone formation at a solid surface as described by Davies (1996). The sequence of

events have initially been observed in vitro (Davies et al., 1991; de Bmijn et ai., 1992).

The events have been confhned by others in vivo (de Bruijn et al., 1995; Mdler-Mai et

al., 1995) at implant surfaces and bone remodeling sites (Zhou et al., 1994). Very specific

interfacial structures have k e n observed at the bone/biomaterial or bonelbone interface

both in vitro and in vivo during de novo bone formation (Davies, 1996). The cernent line,

as descnbed previously, is observed in the early ce11 culture stages, 3-8 days (Davies,

1996). Studies by de Bruijn et ai. (1993) have shown the appearance of these interfacial

structures on hydroxyapatite coated surfaces. In addition, similar interfacial structures

have also been described at the bonelcalcium phosphate based-biomaterial interface in

the in vivo environment (de Lange et al., 1987; de Bruijn et al., 1993)- Figure 1.2

illustrates the stages of new bone formation at a solid surface.

Figure 1.2. Diagrammatic representation of the establishment of an interface between bone tissue and an underlying substrate (Davies, 1 996).

The differentiating bone cells at the substrate surface will secrete a collagen-fiee organic

rnatrix (Fig 1.2A), which provides nucleation sites for the initiation of calcium phosphate

mineralization (Fig. 1 -2B). These non-collagenous proteins act as structural components

in combination with proteoglycans and provide calcium andlor phosphate binding sites

for collagen mineralkation (Davies, 1 996). The calcium phosphate crystals grow and

initial collagen fibre assembly takes place on the organic ma& surface (Fig 1.2C).

Figure 1.2D shows cernent Iine (-0.5 pn thick) whîch thus forms the interface between

the substrate and mineralized collagen cornpartment of bone.

1.A.2. Bone Grafting Materials

1 .A.t.a. Autogenic Materials

Autografts, transplanted material fiom the same i~dividual, are considered to be

the most suitable transplant material primarily because issues of histcornpatr'bility and

nsk of disease transfer from one individual to another are non-existent- There are two

major types of autogenic gr& material: cortical and cancellous. Each type of bone graft

has its associated advantages and disadvantages. The viability of cortical gafk,

specifically, is primarily determined by its ability to revascularize while imparting

mechanical integrity at the defect site. Cortical bone grafts have the potential to produce

good mechanical fïiiing of a defect, although it may take a much longer time to become

viable since either the surface or the interface is the only aspect that becomes completely

revascularized after many years (Habal, 1992). The bulk of the cortical gr& rernains

non-viable for many years but still provides the appropriate mechanical strength (Habal,

1992). Consequently, cortical bone graffs have limited clinical application suice they are

used primarily in areas where there is a need to establish mechanicd integrity such as in

the long bones. Cancellous bone is the choice gr& matenal for achieving fusion and for

correcting discontinuity defects. The autogenic bone is primarily harvested fiom the iliac

bone. Unlike, cortical bone gr&, cancellous grafts do not have the inherent mechanical

strength needed for the reconstruction of large defects, consequently, rigid fixation

devices are required to bridge between the defect area and gr& placement to provide

and/or preserve mechanical strengh to the defect area (Habal, 1992). Such grafts have

large openings, similar to cancellous bone, that allow revascularization to occur, thereby

facilitating new cellular regeneration, remodeling and substitution throughout the gr&

resulting in new bone formation at old bone sites.

Overall, autografts are the most effective material available to augment bone.

However, there are associated limitations when considering them as choice grafts. These

potentiaily include an insunicient amount of @ material needed to augment the defect

site; a significant post-operative nsk of rnorbidity at the donor site (structural weakenhg,

infection, hemorrhage, pain, nerve injury and persistent deformity); and the inability to

mold the autograft in a shape to provide optimal function (Brown et al., 1982; Goldberg

et al., 1993; Mankin et al., 1983; Dick et al., 1985; Makley et al, 1985; Mnaymneth et al.,

1985; Gross et al., 1985). As result of such limitations, an appropriate alternative to

autogr& needs to be considered. Allografts are considered the most cornmonly

emplo y ed alternative.

1 .A.2. b. AUogenic Materials

Lexer (Mnaymneh et al ., 1 985) first used allografts in clinical application in 1 908-

An allograft is bone tissue harvested fiom an individual that is not the recipient.

Allografts generally comprise either cancellous or cortical bone parts and are stored in

bone baaks. The main dserence between autografts and allografts lies in the immuno-

defensive reactions that occur against the latter. Studies have revealed that there is a

distinct pattern of responses associated with allograft implantation. This pattern includes:

(1) the allograft king accepted as an autograft; (2) the rise of irnmuno-defensive

reactions; and (3) the rejection of the allograft due to strong imrnunologic differences

(Fi-iedlaender et ai., 1985 and Horowitz et ai., 1987). Predominantly, the host immune

response is inflammatory and can occur as early as 5 days (Katthagen, 1986). Vascular

ingrowth occurs much more slowly and less extensively than with autogr*. In

addition, these vessels becorne blocked with inflanmatory cells resulting in necrosis.

However, despite these occurrences, a limited amount of oew bone is still formed prior to

necrosis. An accepted ailograil may demonstrate mild callus and repair, although Mted ,

but such repair usually results in nonunion or delayed union and fatigue fractures (Habal,

1992). Consequently, the clinical usefulaess of employing allografts for reconstruction is

questionable and numerous problems associated with using them remain unsolved; such

as the high risk of infection (Hepatitis, AIDS, and bacterial contamination). Furthemore,

the extent of care required today for choosing, checkhg and storing allogenic transplants

makes this technique suniciently expensive that cost supercedes usefulness (Katthagen,

1986).

1.A.2.c. Xenogenic Graf'ts

Another alternative to an autografi is a xenograft. A xenograft is bone harvested

£iom another species that has been subjected to rigorous preparation prior to

implantation. Many methods exist for preparing xenografts, which include fkozen cal f

bone, tieeze-dried calf bone, decalcified ox bone, and deproteinized bone (Hughes et al.,

1943; Hurley et al., 1960; Nade et al., 1977; Salama, 1983). The Kiel bone is one of few

cornrnercially available xenografts prepared by deproteinking bovine caif bone using

hydrogen peroxide as the extracthg agent. It is weakly antigenic but lacks osteoinductive

capability (Heiple et ai., 1967). In fact, in 1970 Schweiberer showed that the Kiel bone

splinter hinders rather than promotes bone regeneration. As a result, the usefulness of the

Kiel bone spiinter is very controversial and only limited success has k e n reported more

recently (Katthagen, 1986). Xenogdb are essentially used as mechanical m e r to

prevent soft tissue ingrowth, which would otherwise iimit osteogenesis. However, to

facilitate osteogensis it is necessary to add components of a u t o m or autologous

marrow to the xenograft (Habal, 1992).

In summary, comparing the digerent bone g r a b available (autogenic. allogeaic

and xenogenic), the autogenic bone graft provides the best results, as there are no

immunlogical problems, its osteogenic capacity is excellent and resorption and

remodeling are quick and effective. However, the limited availability of autograft tissue

and the less than ideal allograft and allograft alternatives have driven the development of

artificial bone-substitute materials. These include the calcium phosphate based-

biornaterials.

1 .A.2.d. Synthetic Bone Substitutes

There has been a shifi in ernphasis placed on the use of gr& material (i-e.

autograh) to synthetic bone substitute implants in the last decade for the reconstruction

of bone defects (Rey, 1998). Particularly, a considerable effort has focused on developing

implant materials composed of calcium phosphates because of theïr close chernical and

crystal resemblance to bone minerai. In fact, calcium phosphate-based materiais have

been employed in medicine and dentistry for over 20 years in such applications that

inchde dental implants, periodontd treatment, alveolar ridge augmentation and

maxillofacial surgery. (de Groot, 1983, 1988; Hulbert et ai., 1987; Jarcho, 1981; Le

Geros, 1988; Le Geros et al., 1993). Consequently, due to the expanding application of

calcium phosphate ceramics, there is a need to investigate various properties that render

them suitable candidate bone-substitutes.

For a calcium phosphate-based ceramic to function successfully as a bone

replacement and or augmenting material it should satisQ certain physicai and biologic

criteria. Initially, if a calcium phosphate ceramic is used in a bone gr& procedure as the

material of choice it should (1) overcome the disadvantages associated with other grafiing

material types (Le- autografis and allografts) but, (2) poses the properties that render

traditional grafting material the superior choice. Consequently, the ideal bone gr&

substitue should: (1) be biocompatible and nonimmunogenic; (2) exhibit osteogenic

properties; that is, the material should actively stimulate the differentiation of

mesenchymai stem cells into active osteoblasts: a property referred to as osteoinductive;

(3) be osteoconductive by providing a ma& for new bone formation; (4) impart

structural strength, both in loaded or stressed sites; (5) be available in uniirnited quantity

and be large enough to be shaped into the size needed; and (6) produce a consistent

biological response that includes its biodegradation during bone healing (LeGeros et al.,

1995).

Several stoichiometries of calcium phosphates (i-e. Caio(P04)6(0H)z - hydroxyapatite, Caio(P04)(F)2 - fluorapatite. Ca3P0& - tricalcium phosphate,

Ca2.7Mgo3(P04)2 - magnesium whitiockite and Ca4(PO4h - tetracalcium phosphate) have

been învestigated for bone repair of which tricalcium phosphate (TCP) and

hydroxyapatite (HA) are the most common. TCPs have k e n formuiated as pastes,

particles, and discs for bone repair (Mors et al., 1975; Ohgushi et ai., 1990a; Nagase et

al., 1991 ; Nagahara et ai., 1992). However. the unpredictable biodegradation profile of

TCP is a troubling issue in bone grafting since TCP biodegradation within bone defects is

routhely not accompanied by bone formation (HoDinger et al, 1996).

Hydroxyapatite is the most extensively d e d calcium phosphate. Heating of the

hydroxyapatite crystals to 2 1 100°C fuses the crystals by the process of sintering and it is

in this form, as a ceramic, that HA has received the most attention. Laboratory derived

HA has biomedical appeal due to apparently sunilar chemistry and in vivo behaviour of

naturai m. Consequedy, an extensive senes of prechhd reports hvolving use for

skeletal applications have been generated. Positive reviews fiom such repom have

demonstrated the suitability of using HA to repair bone (Holmes et al., 1979; Barrows et

al.. 1986; Geesink et al., 1990; Constantine et al., 1992; Frayssinet et ai., 1992; Brown et

al., 1994; Brekke et al., 1998). At present, HA as well as TCP in various forms have FDA

approval for use in bone repair devices and coatïngs on dental and orthopaedic implants.

It is evident, therefore, that calcium phosphates are biocompatible, non-immunogenic and

osteoconductive. Table 1.1 (modified fiom Hollinger et al., 1996) lis& the various HA

and/or TCP-based ceramic products that are distributed in the United States.

Table 1.1 Ceramic Bone Substitutes Distributed ia the United States

Product and Description S ynthograft: TCP Augment: TCP Orthogra!?: TCP Comments: These producrs are particutates and biodegr~dable and should no& be used

Company Johnson and Johnson, Somerville, NJ Miter, Worthington, OH Dupuy, Warsaw, IN

periodontal diseasee Hapset: TCP + calcium sulfate Lifecore Biomedical, Chaska, MN Comments: This product is prepared as a paste for insertion into dental extraction. OsteoGen: Synthetic H A ProOsteon: Coralline-derived HA Comrnents: These products are partieulate

Impladent, Hol~iswood, N'Y Interpore International, Irvine, Ca

and nonbiodegradable and should be used in nonstress-bearing areas, such as periodontaf defections. ProOsteon has been Food and Drug Administration approvedfor use in rneta~hyssal defects. Coilagraft: H A (65%) + TCP(35%) combined with 95% Type 1 bovine collagen and 5% Type III Comments: This pro& is supplied in strips. and the manufacturer suggests that for besr resufts, autogenou &food should be added Application sites recomrnended are spinal_firsium and bone cystic cavities. The collagen and TCP should biodegrade with time.

The principal limitation of calcium phosphate implant materials is their

mechanical properties. These materials are quite brittle, have low impact resistance and

relatively low tensile strength when compared with bone and metais (Table 1.2) (Jarcho,

198 1).

Table 1.2 The Mechanical Properties of Bone, Metallic and Calcium Phosphate Implant Materials

Consequently, this has led to the coating of various rnetals with calcium phosphate in

order to provide the mechanicd properties necessary at the implant site (LeGeros et al.,

Modulus (lo6 psi)

Cortical bone Cancellous bone

MetaIs 3 1 6 L Stainiess Cor-Cr alloy Titanium

Calcium Phosphates Porous Dense

Tensile Strength (10' psi)

Material

Bone

Compressive Strength ( i 3 PS~)

20 6-9

- - -

1-10 30-130

1 0.0 0.5

80- 145 97 50

0.36 10-28

2 -

3 0-40 30 16

5 5-15

1995), while still possessing the advantages of calcium phosphates that include fast bony

adaptation, absence of fibrous tissue, îïrm implant-bone attachent and reduced healing

time (Kay, 1992). However, during high temperature (21500°C) plasma spraying, a

typical coating process, ceramic contaminants, such as nonhydroxyapatites, may be

formed on the surface of the metal prosthesis. These contaminants are chemically less

stable than HA and therefore, could biodegrade, leaving voids throughout the interface,

resulting in the loosening and loss of the prosthesis (LeGeros, 1991). Therefore, to

overcome the poor mechanical properties of calcium phosphates demonstrated at implant

sites, calcium phosphates prepared in a porous form have been investigated. The rationale

for using this form is that the porous form should permit bony ingrowth thereby

reinforcing and stabilizing the implant (de Groot et al., 1 988). Also noteworthy is that the

porosity of a calcium phosphate bone substitute infiuences its biological performance in

vivo, since the rate and distribution of osteogenesis around and throughout the implant

will be effected by size and nuniber of interconnecting channels (LeGeros et al., 1995). It

has been reported that pores of approximately 100 p m in diameter cm provide a

fiamework for bone growth hto the pore volume (Holmes et al., 1988) and becorne easily

vascularized, which is vital to the maintenance of the implant (LeGeros et al., 1995). In

addition, more interconnecting channels can lead to better bone penetration throughout

the bulk material (LeGeros et al., 1995) which enhances the mechanical stability of the

implant at the defect site (Nunes et al., 1997). ProOsteon (interpore International, IMne,

CA, USA) is a FDA approved coral-derived, porous HA ceramic that may be used to

restore nonload bearing metaphyseal defects. HA in this fonn and architecture (500 pm

pore size range and high interconnecting porosity) supports bone ingrowth thereby

pennitting bone-implant stability (Hollinger et al., 1996; Nunes et al., 1997).

Since any practical application (i.e. clinical) of calcium phosphate bioceramics

involves contact with a physiologicai environment., it is important, therefore, to know the

stability or biodegradative potentiai of the implant material. There are essentidy hvo

ways in which a material can degrade, (1) solution mediated and (2) ce11 mediated

processes (Jarcho, 1981). Both processes are believed to be influenced by the

crystal/composition and structure of the material (de Bniijn, 1993). Particdarly, the type

of phase or phases present in the calcium phosphate and the degree of micro- and

macroporosities have significant infiuence in the degradation rate of the material. The

difference in composition and crystallographic structure of such commonly investigated

calcium phosphates such as hydroxyapatite (HA), B-tricalcium phosphate (B-TCP) and a-

tricalcium phosphate (a-TCP) is reflected in the daerence in their stability and

solubility. The order of their relative solubility is a-TCP > P-TCP > HA (LeGeros et al.,

1995). There are different phases of calcium phosphates and depending on their

application, one phase or a combination of phases is used as a potential bone replacement

material. Calcium hydroxyapatite (HA), tricalcium phosphate (TCP) and ratios of the two

OHA/TCP), are the more commonly investigated calcium phosphates for biomedical

application (de Bmïjn, 1993; LeGeros., 1995). HA, Caio(P04)6(OH)z, shows good

biocompatibility when implanted in either soft tissues (Jansen et al., 1985; van

Blitterswijk et al-, 1991; Ogiso et al., 1992) or hard tissue (Jarcho et al., 1977; Denissen

et al,, 1980; de Groot 1981; van Blitterswijk et al., 1985) and has also been shown to

form a strong and intimate bond with bone (Jarcho et al., 1977, 1981; Denissen et al.,

1980). TCP, Ca3(P04)2, has aiso been reported to possess good biocompatible properties

(Nerry et al., 1975, Klein et al., 1983, 199 1 ; van Blitterswijk et al., 1989) but dissolves

rapidly in vivo. (Klein et al., 1983, 1990). CP composites (Le. W C P ) are among the

most widely investigated CP cerarnic biomaterials due to the benefits of their combined

properties (de Bmijn, 1993). These materials, called biphasic calcium phosphate

ceramics, offer several distinct advantages over either phase done- HA is characterized as

relatively bioïnert; that is, it is the more stable phase in physiological envkonment

compared with TCP (de Groot et al, 1992). Consequently, the biphasic materid has the

ability to provide a matnx for new bone growth due to the presence of HA, as well as

biodegrade due to the presence of TCP which is relatively unstable in physiological

solution (LeGeros et al., 1995). Some authors have studied the effectiveness of various

ratios of HA and TCP (Flately et ai., 1983; Berry et al., 1986; Eschenroeder et al., 1987;

Frayssinet et al., 1993). Different combinations of HA and TCP cm be fabricated to

attempt to control the degradation profile of the implant material.

In addition to the importance of solution mediated dissolution in the degradation

of calcium phosphates, there has been evidence published that demonstrates macrophage

and muitinucleated ce11 mediated degradation (van Blitterswijk, et al, 1985, 1989) and

osteoclastic mediated degradation (Davies et al., 1989; Muller-Mai et al., 1990; Daculsi

et al., 1 WO&b, 199 1; Bauer et ai., 1991). Osteoclasts have very precise function and their

activity is critical to the maintenance of the skeletal system, as well as their potential

participation in degrading CP bioceramics. The activity of osteoclasts related to the

overall bone remodeling process can only be rneasured in vivo but the activities of

individual cells is difficult to assess. In the last decade seved groups have introduced

different types of culture systems to study the events occuming during biodegradation of

CP bioceramics (Jones et al., 1984; Ogura, 199 1; Davies et al., 1993; Benahrned et al.,

1994). There is convinchg evidence of osteoclastic resorption in viîro of thin films or

disks of HA (Davies et al., 1992; Fujimonetal et al., 1998) and biphasic HA@-TCP

composites (Davies et al., 1993; Soueidan et ai., 1995). However, there is no evidence

to-date that demonstrates osteoclastic resorption of 3-D prous CP sca.Eolds. Certain

classical morphotogicd and histochemicd characteristics of osteoclasts such as multi-

nuclearity and tartrate resistant acid phosphatase (TRAP) activity may provide supporthg

evidence for osteoclastic phenotype (Davies, 199 1 ), but there are not considered reliable

markers of osteoclastic differentiation (Hattersley et al., 1989).

Currently commercially available synthetic bioceramics can only be used as

filling matenai or as supportïng scaffold without osteogenic capacity. However, to render

such bioceramics osteoinductive, invenigaton have sought to reconstitute the bioceramic

with growth factors or osteogenic proteins that are able to induce or irnprove bone

regeneration. Bone morphogenic protein (BMP) is a glycoprotein present in bone ma&

that plays an important role in both condrogenesis and osteogenesis in embryonic as well

as in post-fetal life (Urist, 1997). Large critical size defects have been successfÙlly healed

by naturaily occwring or recombinant BMP that was carrïed by a suitable delivery

vehicle (Johnson et ai., 1988; Cook et aI., 1994; Cook et al.. 1997). Bioceramics have

been used as carriers of BMP in experhental studies (Lindhoh et al., 1993). Studies

performed by Gao et al., (1996) used a composite bone substitue composed of porous

TCP, BMP and type IV collagen to repair a diaphyseal segmental defect in the tibia of

sheep. A healing superiority was observed when the segmental defect was filled with the

composite bone substitute than with TCP + coiiagen alone. This study showed that the

composite possessed both osteoconductive and osteoinductive properties.

It is evident that there are impressive positive reviews of the use of calcium

phosphate as a suitable bone graft materiai. However, large defects, those resulting fiom

orthopaedic injury or removal of osteosarcornas, requke bone regeneration to occur on a

larger scale where the use of auto/allogenous gdb or synthetic implants are not

satisfactory. Consequently, a new approach to healing bone trauma, called bone tissue

engineering, has emerged to deal with the Limitations of traditional bone transplantation

procedures.

1.B. Tissue Engineering

1.B.1. General

The loss or failure of an organ or tissue is a fiequent, devastating and costly

problem in human health care. For example, every year, millions of Americans suffer

tissue loss or organ failure that result in total national health care costs for patients that

exceed $400 billion per year (Langer et al-, 1993). Arnong these patients, approximately

8 million surgical procedures are preformed annually to treat the disorders (Langer et al.,

1993) which include using implants. In the US alone, about 140 000 artificial hip joint

implants and 20 000 knee prostheses are implanted per year and more than 100 000

patients with relevant defects in joint cartilage are known (Minuth et al., 1998). It was

common up to now to use metal prostheses for replacing hip joints but such materials

usually result in major problems that include implant loosening, inappropriate

degradation, infiammation and blood clotting (Minuth et al., 1998). In order to dirninish

costs of transplantations and pst-surgical complications associated with organ and tissue

repair a new field of study has emerged to provide an alternative solution to tissue

creation and repair. Tissue engineering is an interdisciplinary field that combines the

principles of engineering and the life sciences toward the development of biological

substitutes that aim to restore, maintain, or improve tissue fiction. The concept of

tissue engineering was fust established in 1987 at the US National Science Foundation in

Washington DC. The general strategy that has k e n adopted for the creation of new tissue

involves placing cells on and within matrices. The general strategy involves isolating

cells fiom the body and applying them to a matrk in an in viîru environment. In this

environment, the celis are allowed to grow and differentiate throughout the matrix. The

cells attached to the matrix can then be implanted and become incorporated into the body.

In the body, the matrix can now function as a tissue replacement material and encourage

a faster rate of tissue repair.

l.B.2. Bone Tissue Engineering

At present, bone is the second most transplanted tissue in the USA and Europe

(Martin et al., 1997), consequently, tissue engineering offers a fascinating new alternative

to traditional solutions of bone repair. The underlying concept of bone tissue engineering

involves isolating bone marrow cells, containhg the osteogenic ceii population, fiom a

patient, expanding the population in ce11 culture and seeding them onto a scaf5old. This

materiaVbiologica1 composite, or TE construct could then be grafted back into the same

patient to fûnction as replacement bone tissue.

Advances in ce11 cuIture technology have been fbndamental in establishing the

tissue engineering field since an in viîro phase is key to perfonning the TE strategy.

Currently, bone tissue engineering approaches focus on using mesenchymai stem cells

(MSC) for regeneratïng bony defects MSC have the capacity to differentiate into various

ce11 types that include osteoblasts that give rise to bone (Caplan, 1991). Minuth et al-

(1998) describes three principle steps that must be achieved during the in vitro stage.

Applying these steps to the bone tissue engineering strategy would involve (1) ob-g

sufficient multiplication, proliferation and spreading of mesenchyrnal stem cells (MSC)

on a tissue culture substrate; (2) seeding the expanded ce11 population on a suitable

scaffold and (3) maintainine the differentiated phenotype long-term.

MSCs can differentiate into a number of phenotypes that include bone, cartilage,

tendon/ligament, muscle, marrow and connective tissue, by entering discrete

differentiation pathways. Two different strategies can be employed to achieve MSC-

mediated tissue repair. The first strategy uses the culture-expanded MSCs in their

undifferentiated state. With the presence of local environmental cues, the MSCs wiU then

differentiate into the appropriate ce11 lineage that is responsible for the tissue regeneration

process. The other strategy involves using culture-expanded MSCs that have been

directed ex-vivo into a specific lineage prior to implantation, thus accelerating the healing

process. In the case of bone regeneration, the culture can be directed into the osteogenic

lineage by the addition of various growth factors, including dexamethasone, or cytokines

(Bruder et al., 1994; Jaiswal et al., 1997). Regardless of the strategy chosen, it is essential

that the regulation of MSC proliferation and differentiation be maximized at the in vitro

stage. Maintainhg ce11 differentiation and proliferation is successfdly achieved by

tramferring the tissue carriers into culture containers that are pennanently perfûsed with

fiesh culture medium (Minuth et al., 1992). Continuous elimination of waste products

that would otherwîse be detrimental to cellular fiuiction is alsu achieved by perfusion

(Sittinger et al., 1996).

In addition, the in viîro phase of the TE strategy involves the development of

suitable scaffolds for MSCs seeding and later, in vivo, new tissue support. The ideal

delivery system for osteoprogenitor cells at a bone trauma site wodd be one that niimics

the naturai morphology of bone. The appropnate carrier could allow a three-dimensional

distribution of cells m MCTO thereby accelerating bone healing in vivo. Thus most research

into bone grafting has focused on duplicating the structure and material of trabecular

bone using natural and synthetic structures (Panda et al., 1998; Holy et al., 1998).

l.B.2.a. Macrostructure of Trabecular Bone

The natural open pore geometry of trabecular bone macrostmcture provides a

good starting point for the design of scaffotds for bone tissue engineering. Human

trabecular bone is generally anisotropic, implying that the bone does not appear identical

when held in one position compared to its appearance rotated by 90' (Martin, 1984).

Trabecular anisotrophy is a result of gravitationai stresses imparted by the skeleton. The

upright and bipedal form of human locomotion results in vertical trabeculae being thicker

(200 pm) than horizontal trabeculae (Tobin, 1955). These vertical and horizontal

trabeculae form a highly connected network. Isotropie trabeculae can be found in areas

where the trabeculae are randomly arranged. Figure 1.3 illustrates areas of isotropie and

anisotropic trabeculae that are present in the neck of the human femur.

Figure 13 . Light micrograph showing isotropic and anisotropic areas of the trabeculae in the neck of the femora (Tobin, 1955).

The work of Ward (1838) compartmentalized the types of trabecuiations in the

femora. The medial, M. is located fiom the upper part of the shaft to the articular surface

of the head. The lateral group, L, is situated below the greater trochanter and the upper

surface of the neck. The intertrochantenc arches I l and Ig are shown in Figure 1.4, dong

with M and L.

Figure 1.4. Diagram showing the directions of the trabeculae in the femora (from Instructional Course Lectures, The American Academy of Orthopaedic Surgeons. Vol 10, p.2 15, 1953 .)

The area between L, M and Iz forms a triangular area of structurally weak trabeculatiom

known as Ward's triangle. It was the extensive work of Whiteshouse and Dyson (1974)

and Martin (1984) who described the trabecular bone width and porosity in these

compartments, respectively. Table 1.3 lists the porosity and trabeculae width measured

Table 1.3 Averages of Porosity and Trabeculae Width

1 Area 1 Porosity 1 Ttrrbeculae width

1 Interior of Intertrochantenc / 84.5 I 1.8 1 O. 18 + 0,024

Lateral htertrochanteric arches

- -

79.0 + 5.0 88.2+ 3.2

Arches Greater trochanter

0.23 14 0.053 O. 14 + 0.029

90.51 1 .O 0.3 1k 0.026

Currently, researchers strive to design materiais that have simila. architecture to

trabecular bone in order to mimic the replaced a d o r diseased tissue. Animal studies

have been conducted with the use of porous implants positioned in large defects of long

bones (Hoodendoorn et al., 1984; Daculsi et al., 1990b; Nunes et al., 1997). The results

suggest that there is a f5rm union established at the bone-implant interface. However, the

authors found that f i e r 35 weeks, only one third of the available pore volume was

ingrown with bone. This partîai ingrowth of bone might be attributed to sequestered pores

inside the ceramic; that is, the impiants used had minimal interconnecting pores.

Histological kdings codhmed that bone ingrowth was pronounced near the edges of the

implants in contact with the existing or newly formed bone with no bone growth toward

the centre of the implant (Hoogendoom et ai., 1984). Such implants effectively provide a

scaffold for bone growth but the level of interconnectivity of the implant limits the extent

of bone ingrowth.

1.B2.b. Candidate Scaffolds for Bone Tissue Engineering

Due to the hadequacies of bone gr& that include their heaith related problems

and cost of implementation, there has been a shift in experimental paradigm from the

development of materials to replace or augment bone tissue to materials combined with

cells that can stimulate the body to regenerate tissue more effectively and rapidly. While

bone tissue has been grown in culture for many years, the ideal bone TE construct has yet

to be found. There are various laboratories, worldwide, which are embarking on research

programs directed at bone tissue engineering, with the emphasis on designing the suitable

bone TE constnict-

To asses the potential success and in vivo requirements of a TE scaffold in an

orthotopic site, research groups predominantly use ectopic testing, which involves

implanthg the candidate material into subcutaneous or intramuscular sites (Ohgushi et al,

1989, 1992% b; Goshima et al, 1991% b). Porous calcium phosphate ceramics are the

choice biomaterial used to assay the osteogenic potential of mesenchymal stem cells

(MSC) at ectopic sites primarily because they display excellent osteoconductive

properties. Therefore, the combination of bone marrow derived cells and a calcium

phosphate-based scaffold results in a TE construct that possess both osteogenic and

osteoconductive properties. Results nom such ectopic testing in rats, using a ceramic

composed of 60% HA and 40% TCP having mean pore size of 400 pm, provided by

Zimmer Inc. (Warsaw, lN), combined with rat rnarrow ce11 suspension showed

osteogenesis in the surface pore region at approximately 3 weeks and evidence of

vascular invasion and palisade arrangement of osteoblasts within the pores (Ohgushi et

al., 1 WOa). The ceramic alone subcutaneously implanted, showed no bone formation.

instead, many of the pores were filled with fibrous connective tissue (Ohgushi et al,

1990a). Interpore 200 (Interpore International, bine , CA), as it is referred to in

literature, has also been investigated as a candidate TE constmct. It is produced by

hydrothermal conversion of calcium carbonate skeleton of marine coral (genus Porites).

The resuitant materid is a fiilly intercomected porous HA matrix having an average pore

diameter of 200 W. Both interpore 200 alone and Interpore 200 combined with rat

marrow cells have been implanted in subcutaneous sites in snygeneic rats (Ohgushi, et

al-, 1990a,b, 1992a,b, c; Yoshikawa et al., 1992). Consistent with the resuits using the

Zimmer Inc. material, bone did not form in any implant without marrow cells but bone

did form in all implants with marrow cells after 4 weeks (Ohgushi et al., 1990% 1992%~).

Bone formation occurred initially on the surfaces of the pore w d s . In addition, it was

also reported that bone grew toward the centres of the pores, which was made possible by

the interconnecting porosiq. Other groups have adopted the common TE strategy ushg

biodegradable polymer systems (Ishaug-Riley et al.' 1997; Kadiyala et al., 1997; Holy et

al, 1998) have observed similar results that are seen in ceramic-based TE systems. It is

ciear that bone formation occurs when fiesh marrow is added, but the rate and extent of

healing is a fünction of the amount of marrow and the number of osteoprogenitor cells

residing therein. Consequently, this has led to the use of expanded bone marrow derived

cells as the ce11 population in the in vitro stage. The work of Ishaug-Riley et al. (1 997)

investigated in vivo bone formation on porous (150 - 300 p m and 500 to 710 pm)

biodegradable poly (DL-lactic-CO-gl ycolic acid) foams combined with marrow stromd

osteoblasts (expanded bone marrow denved cells) at an ectopic site in rats. Histological

imaging revealed the formation of mineralized bone-like tissue and capillary invasion in

the polymer/cell composite of the two types of pore size ranges used within 7 days

postimplantation. Uemura et al., (1998) and others (Goshima et al., 1 Wla, b; Yoshikawa

et al., 1996) have demonstrated the strong osteogenic potential of expanded bone marrow

derived cells introduced into macroporous ceramics to fom bone as early as two weeks

in vivo. In contrast, the fiesh marrow/grafl composite showed no significant bone

formation at two weeks. Such results suggest that an expanded ce11 population offers a

greater source of osteogenic cells, which is reflected in the amount of bone formation

observed at earlier time points.

To ultimately test the suitability of a cell/graft composite to reconsmict a bony defect

at an orthotopic site, an appropriate surgicai mode1 must be employed. To implernent the

TE strategy at orthotopic sites in animals, a critical-shed bone defect mode1 is used. The

work of Kadiyala et al. (1997) have demomtrated that a porous material of M C P ,

(Zimmer Inc.), combined with culture-expanded MSC to regenerate a cntical-shed bone

defect in rat femora. Their results revealed that MSC-loaded materials showed

sigdicantly more bone formation at 4 weeks and even more bone formation by 8 weeks

when compared with cell-fkee implants (Kadiyala et al., 1997). However, histological

findings revealed that bone grew preferentidy dong the implantmost interface but not

throughout the bulk material. Similar studies performed by Ohgushi et al. (1992% b, a, c)

reported comparable results with the same material, indicating that the HA/TCP ceramic

corn bined with expanded bone marrow stroma1 cells showed sufficient healing potential

for the treatment of massive bone defects. However, the lack of intercomecting pores

toward the centre of the material resulted in no bone formation in this area- Similar work

has been conducted using a polymer-based TE construct. Puelacher et ai. (1996) placed

PGA-fibre based materials combined with expanded bovine periosteurn-derived cells into

rat femoral sh& defects. Radiographie findings confirrned bone formation at the defect

site on the celvpolymer graft and at 12 weeks, new bone was observed bridging the

surgically created defects completely. In contrast, the animals that received cell-fiee

polymer gr& showed no significant bone formation.

1.B.2.c. Current Limitations of Calcium Phosphate (CP) Bone TE Scaffolds

To successfiilly hc t ion as a bone TE constnict, the candidate biomaterial should

satise certain additional criteria including those mentioned in section 1 .A.2.d. Central to

the TE strategy is designing the scaEold to mimic the lost andor damaged bone. Since

massive bone defects occurring in the long bones are the more challenging defects to fiU

with traditional grafts, a macroporous scaffold that imitates the threeaimensional bone

defect would then be the ideai replacement grafi. The macroporous scaffold would

facilitate bony ingrowth into the pores and subsequently irnpart positional stability, as

discussed in Wolford et al- (1 987). Architecture of the scaffold, therefore, becomes

pivotai in the in vitro stage of the TE strattegy. In the in v&o phase, the engineered

scafFold shouid possess a fiamework for ceii attachent, expansion and differentiation on

and throughout its structure. To accomplish this, the scafEold should have a fully

interconnecting porosity and a pore size diameter that wouid encourage bone ingrowth

into the pore volume- In addition, the appropriate ce11 culture methodology is important

in governing cellular behaviour on the scafTolds. It has k e n demonstrated that scaffold

macroporosity is a critical factor in cell migration and bone matrix elaboration in vitro

(Rout et al, 1987). Matrices with a nominal pore size of 200 Pm (Rout et al, 1987),

resulted in occlusion of pores by migrating ceiis, while similar scaffolds of 500 p

nominal pore size pennitted 3-D tissue growth in vitro (Yoshikawa et al., 1996).

However, Mainard et al. (1996) have reported that a pore size of 80 Fm is necessary for

bone ingrowth in vivo for various calcium phosphates cerarnics (particularly, HA and

HA/TCP systems) implanted into the rabbit femur. Consequently, these results reveal that

the optimum pore size required for bone ingrowth differs in the in vitro and in vivo

environments.

The degradative capacity of a carrier construct is also an important property for TE

applications. As mentioned previously, a biomaterial can be degraded in two ways;

solution mediated and cell-mediated processes. Osteoclastic resorption, a cell-medîated

process, is favored over dissolution because the material becomes replaced by new bone

during the normal course of bone remodeling events in vivo. However, a scafTold that can

support bone ingrowth but also degrade in a marner that coincides with the rate of bone

formation would also be ideal. Polylactide and its derivatives are biodegradable polymers

that have k e n extensively snidied for tissue engineering applications (Holy et al., 1996;

Kadiyala et ai., 1997; Ishaug-Riley et al., 1997). These poIymers can be formed into

three-dimensional structures that mimic the architecture of actual tissue and possess the

appropriate porosity for bone ingrowth. However, despite their capacity to biodegrade by

hydrolysis or by cellular pathways, they lack the mechanical strength to withstand

loading in bone. The commercially available calcium phosphate biomaterials

rnanufactured by Interpore International (Interpore 200 and Pro Osteon 500) fûnction

successfully as bone substitutes. Harvested fiom marine cord exoskeletons, Interpore 200

and Pro Osteon 500 have tme interconnected porosity to ensure complete bone ingrowth.

Through a patented process, a thui calcium phosphate layer is formed on the outer surface

of the calcium carbonate pores and throughout the entire implant. However, despite the

fact that ProOsteon 500 has the appropnate pore size, 500 p, for cell migration in vitro,

the relative importance of dissolution or cell-mediated degradation, and whether this

coincides with the rate of bone formation, is currently unknown for this biocerarnic. As a

result of such Limitations mentioned above, the scaf5olds currently investigated for TE

application do not entirely fùlfill the requirements of a TE constmct since their in viîro

characterization has not been thoroughly investigated.

2. RESEARCH RATIONALE

The underlying concept of bone Tissue Engineering is that bone marrow cells can be

isolated f?om a patient, the ce11 population expanded in ce11 culture and seeded onto a

carrier and the material/biological composite, or TE co~~~ tn i c t , grafted into the same

patient to function as replacement tissue. While bone tissue has been grown in culture for

many years, the ideal carrier for a bone TE constnrct has yet to be found. There are

ceramic-based biornaterials commerciaily available that have been and are currendy

being used in bone tissue engineering research to demonstrate their suitability as TE

consmicts. These materials, however, either lack the interconnecting macroporosity

and/or controiled degradation that make them successfirl bone TE constructs.

The TE strategy comprises an in vitro phase followed by an in vivo phase.

Characterizing the TE scaEold's biological behaviour in vitro has yet to be thoroughly

addressed. The physicai properties (architecture, macro/microporosity, and

intercomecting macroporosity) and the biological behaviour (or the afçiity of the

scaf3old for bone marrow-derived cells) of the TE scafZold in the in vitro phase shouid be

indicative of its success as an ideal TE constnict once grafted into the patient. There is a

need to characteriz a candidate bone TE scafTold in terms of its physical characteristics

so that it satisfies certain defined material properties that render it a suitable TE constnict.

Also, in the in vitro phase the TE scaffold should, through its physical properties, permit

the attachment of bone marrow-derived cells (BMDC) throughout its macroporous

structure, and allow these cells to proliferate, differentiate and elaborate bone matrix. An

important prerequisite is optimal ce11 coverage and attachent be attained through

employing the ideal ceil culture conditions. It is proposed that the ideal conditions to

achieve these optimums would be a dynamic ce11 culture bioreactor that approximates, as

closely as possible, the conditions to which BMDC are exposed in vivo; that is, allowing

continuous flow of medium throughout the porous structure with the goal to accelerate

ce11 proliferation that would otherwise not be attained in static culture conditions.

3. HYPOTHESIS

Intercomecting macroporosity will be the critical factor in detennining the

usefulness of a 3-dimensional calcium phosphate as a bone tissue engineering scaffold in

the in vitro stage of the tissue engineering strategy.

4. OBJECTIVES

4.A. To characterize the macroporosity and intercomecting rnacroporosity of the

calcium phosphates supplied.

4.B. To use different 3-D calcium phosphates, having different macroporosities

and interconnectiag macroporosities, as rat bone marrow ce11 culture substrates in order

to determine whether the distribution of bone fonned in vipo will be a product of their

macrostnicture.

4.C. To develop a method to culture rat bone marrow ceils in 3-D fluid flow

environment to enhance bone matrix elaboration on the cdcium phosphate scaEolds

provided.

5. MA'IERLALS & METIIODS

S.A. Methods of Characterization

S.A. 1. Provision of Cakium Phosphate Samples

Five different types of 3-dimensional porous calcium phosphates were

investigated in this thesis for their suitability as tissue engineering construcl. They were

generously provided by three different sources and are referred to as Zimmer (provided

by Zirnrner International Inc, Warsaw, IN, USA), CAM40/60, CAM70130, 2CAM70/30

(received fiom CAM implants, The Netherlands) and CPP (-60ppi, -45ppi, -20ppi and - 1 Op pi) (O btained fkom the Sc ho01 of Materials Engineering, Y eungnam University,

Kyongsan, Kyongbuk, Korea). The 'ppi' extensions used to descnbe the CPP ceramics

describe the pore size in units of 'pore per inch'. The macroporous CPP scafTolds were

made using polyurethane (PU) sponge method. As part of the procedure, the PU is bumt

out and the resultant inorganic macroporous scaffold that is created has its pore diarneter

defined by the PU sponge diarneter, hence the use of 'pore per inch' units (Baksh et al.,

1998). The rnethodç of processing the Zirnmer and CAM Implant ceramics samples are

not described here due to confidentiality agreements.

5.A.2. Light Photography

Calcium phosphate samples were cut, ushg a diamond ciisc with a dental

handpiece, into blocks having dimensions of approxirnately 4 x 4 x 4 mm and were

viewed at various magnifications under a dissecting microscope (Wild M3Z, Type-S,

Heerbrugg, Switzerland). %y varying the light source, provided by an Intralux@ 5000

light box, over the sample surface, the 3-D configuration of the material was viewed as

well as debris formed during the cutting process. The excess debns was removed using

an Easy Dusterm spray (SPI Supplies, West Chester* PA, USA). Black & white

photographs of the CP blocks at 10, 15, 20 and 40x magnification were generated to

record the macroporosity and extent of interconnecting macroporosity of the calcium

phosphates supplied.

5.A.3. Powder X-Ray Diffraction Spectroscopy (XRD)

Powder XRD was used to confirm the identity, the cxystallinity and phase purity

of the solid CP biomaterials. A fülly automated Siemens D5000 Difiactometer System

was used for data collection. The system operated with Cu-ka radiation on 50kv/35 mA.

A Kevex SS Detector monochromatized the secondary bearn. A step mode was used

during the data collection with step size: 0.02O 2-theta and counting time of 2.0 S. The

scanning range was 0" - 55" (2-theta). Some samples were run 2-3 times to obtain

satisfactory sets. XRD diffiactograms provide information on phases present, phase

concentrations, amorphous content and crystallite size. Software supplied by the

manufacture of the Siemens D5000 Difiactometer System was employed to measure the

relative peak heights using a least squares method to obtain a quantitative estimate of

phase abundance.

5.A.4. Scanning Electron Microscopy (SEM)

As-received CP specimens were prepared for SEM irnaging to observe the extent

of the macro/microporosity and crystal morphology characteristic to them at high

magnification. Dry, as-received CP samples were mounted on specimen stubs using

adhesive glue. Approximately 3 nm layer of platinum was sputter coated with a Polaron

SC515 SEM Coating System onto the CP specimens which were then examhed at

various magnifications in a Hitachi S-2000 scanning electron microscope at an

accelerating voltage of 15 kV. The images generated were used to measure the

rnarco/micropore size range and intercomectivity characteristic of each CP type supplied.

S.A.5. Back Scattered Electron Microscopy @SE)

5.A.S.a. LR White Embedding Protocois

As-received CP samples, of each type supplied, were prepared for LRW

embedding. The specimens were kept in 7096, 80%, 90% aad 95% ethanol for 1 hour and

then immersed in 100% ethanol for 3 hours. A 1 : 1 mixture of LRW (SPI Supplies, West

Chester, PA, USA) and 100% ethanol was prepared to begin the LRW infiltrahg

process. Specimens were placed in this mixture for 1 hour and then subsequentiy placed

in 100% LRW for 1 week under vacuum. The final stage of the embedding process was

initiated by polymerizing the LRW resin by adding I drop of activator per 10 ml of LRW

resin. Sarnples were placed in rnoulds and then filled with activated LRW. This

configuration was kept in a bath of ice-cold water during the embedding process in order

to disperse the heat produced during exothennic polymerization of the resin.

Polymerization occurred in 10-20 minutes, after which the samples, having approxhate

dimension of 4 x 4 x 4 mm, were cut in half using an IsometTM Plus precision saw

(Buehler Ltd., Lake Bluff, II, USA). One of the halves was mounted on a SEM stub and

the exposed surface was polished manually using an EcometB 5 two speed g ~ d e r -

polisher (Buehler Ltd., Lake Bluff, Il, USA) with decreasing grit size paper (120, 240,

400,600, 1200 USA grit size). The polished surface was coated with approximately 1 nm

of plathum and backscattered electron images were generated. Following andysis, the

sampIes were stored in a sealed container lined with dry-rite.

5.A.S.b. Quantification of Macroporosity and Intercoanectivity

Polished LRW embedded as-received CP samples were used to quantiQ the

macroporosity and interconnectivity characteristic of each type of CP supplied.

Backscattered electron images were generated fiom the polished surfaces. However, the

entire surface area could not be imaged even at the lowest attainable magnification on

SEM (~35). Consequendy, various images were captured and montaged in Corel Draw

5.0 to reconstnict the entire surface area Using an image analyzing program (Sigmascan

Pro) the pixel intensity measured fiom the CP material, LRW resin and voids (areas not

filled with LRW) were recorded. The relationship between these intensities provided

measurements of macroporosity and interconnectivity.

5.A.6. Degradation Behaviour of CP Scaffolds

Twenty-four samples of each type of CP supplied were placed in 0.1 M Tris

BufYer (Appendix A) for 6 weeks to assess the degradation of the CP biomaterials in an

acellular environment. Samples were y-sterilized at 2.5 Mrads prior to experimentation.

For each week, up to 6 weeks, the pH of the buffer solution was measured using a pH

meter (pH meter 10 accumet@, Fisher Scientific). In addition, the calcium ion ([ca27)

concentration was determined by atomic absorption spectroscopy.

A CP sarnple of each type was prepared for SEM at each time point studied (1,4,

and 6 weeks) to observe changes in surface morphology as a result of degradation. For

preparing wet samples for SEM imaging, specimens were dehydrated in a graded series

of ethanol and cntical point dried from carbon dioxide in a Polaron CPD7501 Critical

Point Dryer. Kepresentative samples fiom each week were mounted on specimen stubs

and platinum coated for viewing in a scanning electron microscope .

5.B- Static In V i o Studies

S.B.L. Rat Bone Marrow Culture System

Primary bone marrow cultures have the potentid in shoa time periods to give rise

to a population of differentiating osteogenic cells that produce matrix that is

morphologicaily and biochemically distinguished as bone matrix (Maniatopoulos et al.,

1988). This in vitro technique with its cntical dependence on the differentiating cells at

the nibstrate surface can provide information on the formation of bone on synthetic

biomaterial surfaces and a characterization of the structures that are present at the

bonehiomaterial intedice. The description provided betow outlines the Rat Bone

Marrow Protocol developed by Maniatopoulos et al., 1988 and reviewed in Davies, 1996

that was adopted for this research work. Modifications to the protocol were made in order

to accommodate 3-D substrates.

S.B.2. Description of Static CeU Culture Configuration

5.B.2.a. Primary CeU Culture, Subculture & Initial CeU Seeding

At an animal facility, rat femora of a young adult male Wistar rat (approximately

120g) were aseptically removed and dissected fiee of adhering soft tissues following

cervical dislocation and placed in tubes containing 10X antibiotic (AB) solution. The

securely capped tubes containing the rat femora were transferred to a lamina flow hood

at a culture facility. The cleaned femora were soalced in 3 additional 10X AB washes, 10

minutes each time and then briefly washed in a tube containing 10 ml a-Minimal

Essential Medium (a-MEM). The epiphyses of the fernur were cut off to expose the

marrow cavity. Using a 10 ml syringe filled with culture medium, the marrow contents

were emptied into a 15 ml tube via a sterile 20G11/2 needle. The marrow contents were

thoroughly flushed out once the femur tumed white. This procedure was repeated for the

other femur. The harvested rat bone marrow ce11 suspension fiom a rat was transferred

into a sterile 50 ml conical tube. The final volume was adjusted to 30 ml and

supplemented with fieshly prepared ascorbic acid (AA) (50pg/ml), B-glycerophosphate

(BGP) (5 mM) and dexarnethasone @EX) (10-~ hQM). A 15 ml volume was pipetted in a

T75 flask and placed at 37°C in a hurnidified COz incubator (5% CO2 and 95% air). The

culture medium mas changed every 24 hours. This procedure was repeated for ai l

primary cultures.

On Day 5 of the culture, the prîmary rat bone marrow cultures was subcultured

using a trypsinization protocol that removes adherent cells. The prirnary culture medium

was aspirated; the culture rinsed twice with 10 ml phosphate buffered saiine (Ca-Mg fiee,

PBS-CaMg) each time. The b s e r was aspirated and 5 ml 0.01% trypsin was added to

the culture and the culture was incubated for 20 minutes at 37°C. The culture flask

contents were pipetted up and down several tirnes to release the adherent cells and then

washed with 5 mi senun contaullng medium to neutralize the trypsin. The ce11

suspension was pipetted in to a centrifuge tube (Blue Maxm 50 ml polypropylene conical

tube, Becton Dickson Labware) and spun down at 80-1 00g in a centrifbge (Omnifuge

RT, Canlab) for 10 minutes. The supernatant was aspirated and the pellet r emaihg was

resuspended with 10 ml of culture medium. The suspension was filtered through a 100

p m ce11 strainer (Cell strainer 100 pm Nylon sterildgamma-irradiated, Becton Dickson

Labware). The ce11 count was detemiioed using a Couiter counter. The maximum celi

seeding densities were prepared (Appendix B).

CP scafTolds were placed in the wells of a sterile 24-well microplate. A total

volume of 2 ml of a-MEM was placed in each well containiog the CP sample for 10

minutes. Prior to ce11 seeciing, the a-MEM solution was aspirated fiom each well. Using

the desired ce11 seeduig concentration, 1 ml of this solution was seeded on to the top

surface of each sample using a 10 ml syrhge fked with a 20G11/2 precision glide

needle? positioned at the centre of the sample (Figure 5.1).

24-well plate 3-D substrate with needIe positioned at centre

Figure 5.1. Ce11 Seeding on to 3-D Substrate

The culture was placed in an incubator mainbined at 37*C, 95% air and 5% COz for the

desired ce11 seeding tirne. After ceil seeding, each sample was transferred to a 15 ml

sterile po ly styrene round-bottom tube and replenished with fully supplemented culture

medium. This configuration was placed in the same incubator and re-fed with growth

medium every 24 hours until tennination. At termination, the well contents were rinsed

&ce in PBS-CaMg and a-MEM, respectively, prior to king fixed in Karnovsky's

fixative (Appendk C) ovemight. M e r ceii £kation, the samples were stored in 0.1 M

sodium cacodylate buffer (pH 7.4,4OC) until M e r analysis was performed.

Prior to al1 ceil cultures, CP samples were cut using a diamond disc-fitted dental

handpiece drill into blocks havhg approximate dimension of 4 x 4 x 4 mm and sterilized

by y-irradiation (2.5 Mrads).

5.B.2.b. Tissue Culture Substrate for Celi Seeding

To maximize ce11 adherence to the CP sampIes. the appropriate tissue culture

substrate needs to be empioyed. To determine the optimum substrate, two types of

culture substrates were use& tissue culture treated (TCT) and bacteriological-grade (BG)

24-well plates. The desired ce11 seeding density was prepared fiom 5-day primary bone

ce11 cultures, as described in Appendix B. Representative CP specimens were placed in

wells of both TCT and BG 24-we11 plates and soaked in a-MEM for 10 minutes. Prior to

ce11 seeding, the a-MEM solution was aspirated fiom each well. Samples were seeded

with the prepared seeding density for 1 hour. One well contained no sample as a control.

This cell configuration was placed in au incubator maintained at 37OC and 5% COz for 2

weeks. The ce11 culture was re-feed with tùlly supplemented medium every 24 hours

until tennination. At tennination, the samples were £ked in Karnovsky's fixative and

then prepared for SEM analysis. This experiment was repeated two additional times to

confïrm consistency in observations,

5.B.2.c. Ceil Adberence as a function of Seeding Time

To determine an appropriate ce11 seeding tirne, primary RBM cells were

subcultured on Day 5 and using the maximum ce11 seeding density, the cells were seeded

into wells of a 24-well microplate. The cultures were kept at 3?C, 95% air and 5% CO2

for 0.5, 1, 3 and 4 hours. At each time point, the media was aspirated fiom 3 wells and

the contents of the wells were trypinsinzed for 30 minutes. The cells £iom the aspirated

media (non-adherent cells) and the ûypinsinzed cells (adherent cells) were counted in a

Coulter counter. The results were plotted and the appropriate ce11 seeding time was

determined. This ce11 seeding time would be used in subsequent cultures involving 3-D

substrates.

5.B.2.d- CeU Adherence on 3-D substrates during Cell S d i o g

To assess ce11 colonization on 3-D substrates at the chosen cell seeding tirne, four

samples of the Zimmer-type, CAM Implant-type, and CPP-type were seeded with the

maximum ce11 seeding density attainable at subculture. Tissue culture treated substrates

were used as positive controls and bacteriological grade substrates were used as negative

controls. Each sarnple type was placed in the well of a BG 24-well plate. The cellç were

seeded, as described previously, for 1 hour. The side opposite the seeding surface of the

block piece was marked using a permanent marker so events only occming on the

seeding surface could easily be interpreted. After ce11 seeding, the culture was incubated

at 37°C in hurnidified atmosphere of 95% air and 5% CO* for the chosen ce11 seeding

time. Once the seeding time was anained, three samples of each CP type were placed into

a new well, thoroughly washed in PBS-CaMg for 10 minutes, trypsinized for 1 hour, and

a ce11 number was detennined using a Coulter counter. Following trysinization a sample

of each type was stained with Toluidine blue and viewed in a dissecting microscope to

observe the presence of remnant tissue componets. The wells containhg the samples

were also trypsinized for 1 hour. The ce11 quantity attached to the well plate and in the

medium after 1 hour was also counted. In addition, a CP sample was fixed with

Karnovslcy's fixative and prepared for SEM anaiysis to observe the morphology of the

cells colonizing the surface of the CP. AU critical point dried samples were mounted on

SEM stubs with the marked side of the block glued to the surface of the stub.

S.C. Dexamethasone (-) Cultures

To assess the osteoclastic resorptive potential of the CP samples supplieci,

specimens were placed in ce11 culture medium containing no dexamethasone (DEX-).

Bone marrow cells were harvested fkom rat fernom Four samples, each placed in the well

of a BG 24-well plate, was seeded with 2 ml of RBMD cells following the method

described previously. The cultures were incubated for 1 hour at 37OC in 5% CO2 and 95%

air, subsequentiy, each CP samples was placed at the bottom of 15 ml polystyrene round-

bottom tubes and replenished with DEX- culture medium. This configuration was

incubated for 4 weeks. At 1,2,3 and 4 weeks, a sarnple was rinsed in PBS-CaMg and a-

MEM three times, 10 minutes each, respectively, pior to being placed in Kamovsky's

fixative overnight. On the next day, samples were thoroughly washed in 0.1 M sodium

cacodylate buffer and prepared for SEM imaging.

5.D. Dynamic In Ufro Studies

5.D.1. Description of Dynamic Ce11 Culture Configuration

Bone marrow cells were harvested fiom rat femora and placed in the ce11 culture

system descnbed previously. On Day 5 the primary culture was subcultured and the

desired ce11 seeding density was prepared. Gamma-sterilized CP samples were placed in

the weiis of BG 24-well plates and washed with a-MEM for 10 minutes. The ce11

suspension was then seeded on to the top surface of a sarnple using a 20G11/2 needle.

This configuration was placed in an incubator maintained at 37OC and 5% COz for 1 hour.

At 1 hour, the sample was transferred to a sterile 50 ml polystyrene conical tube filled

with 40 ml of M y supplemented medium containing long-acting ascorbic acid

(Appendix D). The conical tube was tightened using a screw cap and placed in a 3-D

medium-rotating configuration. Cultures were incubated at 37OC in 100% air (or 0%

COî). All cultures placed in this system were re-fed weekly until termination.

The rotating system employed contaios four glass tubes; open at one end,

positioned in a manner that achieves 3-D flow of culture medium- Each g las tube was

designed to hold two 50 ml conical tubes. Each glass tube was sealed using a rubber

stopper. A motor, operating at 120 Volts and 0.10 Amps, k e d to the central shaft

holding the specimen modules, allowed the system to be rotated continuously at 50/60

cylces/s for the duration of the culture. The sample placed in each tube was stabüized

between stainless steel mesh wires cut to the same diameter as the tube. Figure 5.2

illustrates the design configuration

Figure 5.2. Dynamic Cell Culture Apparatus

S.E. Cell Colonïzation, Arrangement and Function in Static & Dynamic Culture Systems as a hct ion of Time

Calcium phosphates specimens were placed in static and dynamic culture

environments for 1,2,3,4,5, and 6 weeks to study bone tissue growth on and throughout

the their 3-D structure. For each week, up to 6 weeks, samples were washed in PBS-

CaMg, a-MEM and fixed in Kaniovsky's fixative overnight, At 1, 4 and 6 weeks.

samples were prepared for SEM and histology.

5.F. In Vwo Studies

5.F. 1. Surgical Procedures

Calcium polyphosphate scafKolds (n=8) were y-irradiated and implanted

aseptically in male Wistar rat (200-250g) femora under 4% haiothane, nitrous oxide and

oxygen (2: 1) anaesthesia. The implants were press-fitted into tram-femoral drill holes

made using a low speed dental bandpiece and 2 mm round bwr, and the incision was

closed using 4-0 polygylcolic acid sutures and skin staples. Each animal (n=4) received

one implant per femur and was allowed fidl activity postoperatively. In vivo implants

were fixed, after 1, 2, 6 and 23 weeks, by immersion in Karnovsky's fixative for 24

hours. For the rats sacrificed after 1 and 2 weeks, the rat femora were prepared for SEM

imaging and the left femora were processed for histology. Both femora of the rats

sacrificed at 6 and 23 weeks were prepared for histology.

S.F.2. Histological Preparations

S.F.2.a. In Vivo Samples

Excised rat femora were tnmmed to include the portion of the diaphysis

containhg the implant and immersed in 10% buffered Formalin (pH 7.2) ovemight.

Fixed samples were then placed in 14.4% EDTA in 10% Formalin (pH 7.0 - 7.2) for

approximately 4 - 5 weeks. The solution was changed every second day and the

decalcieing bone was kept on a Red Rotor rotating table (Hoefer Scientific Instruments,

San Fransico, USA). Decalcification was c o b e d by X-ray radiography of the

diaphysis containhg the implant- The femora containing the implant was lefi in ninnuig

water overnight to neutralize the decalcïfjring action of the formic acid. Decalcified

sarnples were then dehydrated through a graded series of aicohols. The final stages of the

dehydration process involved placing the femora in 1:l ratios of 10W0

methylbenzoate/lOO% ethanol and nnally, 100% methylbenzoate until the bone appeared

transparent. The femora were then infiltrated with xylene for 1 hours and then with

xylenelparafnn for 2 hours, followed by pure paratnn (surgiPath@ Blue Ribbon)

idiltration under vacuum at 59'C. Once the paraffin had hardened, thin sections (4 pm)

were cut usine an American Optical microtome (mode1 820) with a cutting angle set at

6.5'- The cut paraffin was floated in a heated bath of water (46 - 4g°C). The ribbons were

carefully mounted on acid clean glass slides (VWR CanLab, 1" x 3") then deparaffinized

3 times for 5 minutes in xylene, and rehydrated in decreasing concentrations of alcohol to

water prior to staining. SLides were then stained in Harris' Hematoxyiin (surgiPath@) for

5 minutes and counter-stained in Aqueous Eosin. A second series of ethanol washes were

performed, this t h e reversing the solution concentration back to 100%, finishing with a

wash in xylene. Coverslips were placed on the slides and mounted with Entellan

mounting media (Merck). Histological sections were examined by light microscopy

(Leitz, Heerbrugg, Switzerland) and colour pictures were taken using Ektachrome T64

coloured reversal film.

5.F.2.b. In Vitro Samples

Fixed calcium phosphate samples fkom in vitro experiments were infïitrated with

molten agar (2% aq), The agar solidified around and throughout the sample at room

temperature. The inNtrated samples were placed in 10% Formalin overnight to harden

the agar. On the next &y, this complex was placed in 14.4% EDTA in 10% Formalin

until the CP sample was decalcified, as confumed by X-ray radiography. Decalcification

took anywhere fkom 5-8 weeks depending on the sample type. Each mould containhg a

sample was processed for histology as described abve. Serial sections were made

through half the sample in order to observe the extent of bone growth throughout the

sample. The halfway point corresponding to approximately 2 nim into the sampie (total

length of the samples were approdately 4 mm). Histological sections obtained fiom the

halfway point of each sample were examùied by light microscopy and colour pictures

were taken of these sections.

6. RESULTS

6.A. Methods of Characterization

6.A.1. Light Photography of As-Received Samples

The macroscopic structurai appearance of the 3-D calcium phosphates supplied

was clearly visualized by Light photography, as shown in Figures 6.1A - 6.8A (following

pages). It was evident, at this mapification, h î there was a varying degree of

rnacroporosity associated between each CP type. Specificaily, the distribution and size of

the macropores throughout the CPP types varied as seen in Figures 6.5 - 6.8. At higher

magnificatio~ the ievel of interconnecthg macroporosity was obsewed, as shown in

Figures 6.1B - 6.8B. It is evident that the greatest degree of interconnections between

pores was associated with the CPP sampte types. Also apparent was that the PU sponge

method used to create the CPP material was successful in creating marcoporous scaffolds

of varying pore sizes as seen when comparing Figures 6.1 - 6.8 (following pages). The

Zimmer and CAM Implant samples appeared to have similar surface appearances. Their

surfaces were pitted with pore openings that showed minimaai connections between

neighboring surface pores. Varying the light over these samples reveded that the samples

had a certain degree of porosity associated with them since Iight, as seen wi?h the naked

eye, penetrated through the bulk of the sample.

The samples depicted in the light photographs were typical samples used in future

experiments. All samples were cleaned of excess debris with an Easy Dusterm and 6 -

sterilized (2.5 Mrads). The physical appearance (Le. coiour) of the samples was not

effected by the sterilization procedure.

Figures 6.1A - B. Light photographs showing the Zimmer d o l d as-ieceived. Field widîh = 24 mm for A and field width = 12 mm for B.

Fi yres 6.2A - B. Light photographs showing the CAM40/60 scaBold as-received. Field width = 24 mm for A and field width = 12 mm for B.

Figures 6.3A - B. Light photographs showing the CAM70f30 scaffold as-received. Field width = 24 mm for A and field width = 19 mm for B.

Figures 6.4A - B. Light photographs showing the 2CAM70/30 scaffold as-received. Field width = 24 mm for A and field width = 19 mm for B.

Figures 6.5A - B. Light photographs showing the CPPdûppi scaffold as-received. Field width = 24 mm for A and fieid width = 19 mm for B.

Figures 6.6A - B. Light photographs showing the CPP4Sppi scafZold as-received. Field width = 0.54 mm for A and field width = 0.27 mm for B.

Figures 6.7A - B. Light photographs showing the CPP-20ppi scaffold as-received. Field width = 0.54 mm for A and field width = 0.27 mm for B.

Figures 6.8A - B. Light photographs showing the CPP-1 Oppi scaffold as-received. Field width = 0.54 mm for A and field width = 0.27 mm for B.

6.A.2 Powder X-Ray Diffkaction Spectroscopy (XRD)

Pulverized as-received CP samples were analyzed utiiizîng an X-ray

difhctorneter to confinn identity, presence of crystaliine/amorphous phases and phase

purity. The results are presented in Table 6.1.

Table 6.1 XRD Results of as-received Calcium Phosphate ScafTolds

As-received CP 1 Crystabity Zimmer

CAM40/60

Crystalline h

Crystalline

Cry stalline

XAM70/30 Crystalline

The XRD spectra for the various samples are shown in Graphs 6.1 - 6.5 (following pages).

The experimental ciiffiaction patterns were compared with calculated ones based on the

structural data for HA, TCP (a and types) and P-CMP available in the ICDD-Database

(International Centre for Diffraction Data) (1998). Powder x-ray difiction patterns were

generated from as-received calcium phosphates fiom the different manufacturers

(Zimmer, CAM Implants, Sc ho01 of Materials Engineering, Yeungnam University,

Phases 2-phase mixture of HA and f!-TCP 2-phase system of HA and B-TCP 3-phase mixture of HA, and two polyrnorphous of TCP: P-TCP and a- TCP 2 component system of HA and a-TCP B - P ~ ~ Y ~ Y P ~

Note: This compound is known ro have 3 polymorphous: alpha, beta and gamma. They have d~xerent structures and d~pac t ion patterns.

Phase Purïty 60% HA

,40% W C P 40% HA 60% B-TCP 70% HA 30% TCP

70% HA 30% a-TCP Pure B-CMP (Ca- metaphosphate, ca(?o3)2

Korea) labeled (A) CAM40/60; (B) CAM70/30; (C) Zimmer, @) CPP and (E)

2CAM70/30. The XRD patterns revealed crystallinity and variable compositions. The

black arrows on (A), (B) and (C) correspond to the peaks characteristic to f3-TCP while

the remaining peaks indicate the presence of HA. The peaks marked with asterisks on (B)

and (E) indicate the presence o f a-TCP. The amount of a-TCP (wt %) found in (E)

2CAM70/30 was higher than in (B) CAM70/30. h p h @) shows the CPP sample that

corresponds to the XRD spectrum of Ca(P03)z, CMP (calcium meta-phosphate).

/ Graph 6.1 Powder XRD pattern of CAM40/60 !

1 - - - --- -- -- __ _. --- 1

Graph 6.2 Powder XRD of CAM70/30 1 i

Gmph -- - -

10 14 18 22 26 30 34 38 42 46 50 54

2 theta

6.3 Powder XRD of Zimmer sample ----__ -- -- - - - - - - - - -

Graph 6.5 Powder XRD of 2CAM70/30

6.A.3. Scanning Electron Microscopy

6.A3.a Micro and Macroporosity

Scanning electron microscopy was used to observe the morphological appearance

of the components that comprise the CP samples and the extent of macro/microporosi~

characteristic of each CP-type. Figures 6.9A - M and 6.10A - 1 are micrographs of

uicreasing magnification illustrating the surface morphology of the various CP materiais.

AU the CP samples exhibited a d a c e morphology containing both macro and

microporosity, as seen at low magnification (Figures 6.9A7 E, 1 and M and Figures 6. lOA,

DI and G) and at higher magnification (Figures 6.9D, H, L and P and Figures C, F and I),

respectively. The processing procedure used to create the porous scafSolds resulted in

fused calcium phosphate grains as seen in Figures 6.9C7 G, K and O and Figures 6.1 1A,

E and H that resulted in the microporosity characteristic of each CP type. The calcium

phosphate grains that comprise each ceramic type were clearly visualized at this

magnification (x5.OK) and it was evident that the size, shape and morphology of the

grains differ between CP types- Both the Zimrner and CAM Implant samples

demonstrated a heterogeneous surface. The Zimrner-type material constituted irregularly

shaped grains that were fused together. While the CPP materiai comprised cylindrically

shaped grains of varying sizes that were also fused together. Within the CAM Implant

sample types (CAM40/60, CAM70130 and 2CAM70/30), the calcium phosphate grains

appeared spherical. However, the 2CAM70130 sample had its microporosity created by

staggered leaflet structures positioned between spherical grains.

Figures 6.9A - P. Scanning electron micrographs of as-received Zimmer and CAM Implant scafEolds. Micrographs A- D, E - H, 1 - L, M - P, correspond to increasing magnification of the surfaces o f the as-received Zimmer, CAM40/60, CAM70/30, and 2CAM70/30 porous ceramics. The lower magnification micrographs (A, E, I and M) demonstrate the level o f macroporosity, while the individuai grains and microporosity are seen at higher magnifications. F.W. = 2.58 mm, 90 Fm, 18 pn and 9 pm for A, B, C and D, respectively. F.W. = 1.8 mm, 90 pm, 18 p m and 6 pn for E, F, G and H, respectively. F.W. = 2.25 mm, 90 pm, 36 pm and 18 pm for 1, J, K and L, respectively. F.W. = 2.25 mm, 90 pm, 18 pm and 6 p for M, N, O and P, respectively.

Figure 6.10A - 1 Scanning electron micrographs of as-received CPP samples. Labels A - C, D - F and G - 1 correspond to increasing magnifications of the CPP-45ppi, CPP-Zûppi and CPP-lûppi sample typcs, respectively. Micrographs A, B and C have increasing magnincations of x35, x1.00K and x5.OOK corresponding to field widths of 2.58 mm, 90 p and 18 p.m. The macropore size range and level of interconnectivity is clearly seen in A, D and G while in C, F and 1 the individuai grains and micropores comprising the CPP types are depicted.

SEM was also used to estimate the grain size, micro/macropore size range

characteristic of each type of CP. This method of pore size characterization was chosen in

preference to other methods such as mercury porosimetry because 1) it gave a direct

representation of the surface porosity available for bone groowth and 2) mercury

porosimetry is restricted to size evaluation of h e r sized pores (> 20 um) (Cameron et al.,

1976). Table 6.2 lists the micro/macrostructuraI properties of the porous ceramics.

Table 6.2 Micro/macrostructural Properties of the as-received scafEoIds

Maeropore size range

Ceramic

Zimmer

6.A.3.b. Interconnectivity

2CAM70/30 CPP-45ppi CPP-2Oppi CPP- 1 Oppi

Figures 6.10A - I show three different pore size ranges manufactured using the

PU sponge method (45ppi, 20ppi, l0ppi). At low mapification, (x35, Figures 6.10A, D

Average Grain Size [clml

and G), the continuous macroporosity in al1 three orthogonal directions is shown. This

Micropore sue range

7.8 7 7 7

high degree of pore continuity was not observed in the Zimmer and CAM Implant

[clml

scaffolds. The pore structure of these CPP scaf5olds was not uniform throughout the

[ P l 1.7 1 0.6 - 1-7

0.4 - 0.8 1 50 - 500

matrix. This observation was consistent with the Zimmer and CAM Implant samples.

50 - 400

2-5 1-2 2 - 7

W i h each sample-type, there were pore openings that appeared both round and

elongated. The intercomecting channels of the CPP-45ppi, CPP-20ppi and CPP- 1 Oppi

450 - 720 855 - 1335 1000 - 1600

were measured to be 325, 580 and 1000 pn on average, respectively. When comparing

the CPP matrix to human cancellous bone harvested fiom the femoral neck, at similar

magnification (Figure 6.1 l), it was clear that the honeywmb structure characteristic of

trabecuiar bone was similar to that of the CPP structures.

Figure 6.11. Scanning electron micrograph of human trabecdar bone. The variation macropore size and extent of intenwmecting macroporosity is clearly evident at this magnification (x35K). Field width = 2.58 mm.

The CPP materials demonstrateci full intercomecting macroporosity that was also similar

to that typical of trabecular bone. In fact, CPP-45ppi and CPP-2ûppi samples appeared to

have distinctively similar pore sue ranges and interconnecting channels when compared

to trabecular bone. Consequently, it was because of these obsewations that the CPP-

45ppi and CPP-20ppi types were used in preference to the CPP-lOppi type in fitute

experiments. The Zimmer and CAM Implant ceramics did not reveal macrostmctural

simiiarities to human trabecdar bone.

6.A.4. Quantification of Total Porosity

Backscattered electron images were used to quant* total p o r o s i ~ that included

both micro and macroporosity of the CP materials supplied. In order to generate such

images in a scanning electron microscope, the samples were infïltrated with LRW resin,

polished and coated with a thin layer of piatinum- The images captured ~epresented the

surface area midway through the sample. At 50x magnification random spots were

imaged for each sample type. Al1 imaged areas measured 8.9 cm x 11.5 cm. Four images

were generated from three different samples of the same matenal type. This was repeated

for al1 calcium phosphates supplied.

Using Sigmascan Pro, an image-anaiyzing program, the pixel intensity associated

with the CP matenal, LRW resin and void areas were measured. On a backscattered

image, the CP material, LRW resin and void areas appeared white, gray and black,

respectively, as seen in Figure 6.12.

Figure 6.12. A typical back-scattered electron @SE) Mage generated from a LRW embedded CP scaffold. The BSE image shows a 2CAM70/30 sample infiltrated with resin. Similar BSE images were generated for Zimmer, CAM40/60 and CAM70/30 sample types. This BSE image shows areas of white, black and dark grey that correspond to the CP material, LRW resin and void areas, respectively. The void areas represent both micro and macropores. Cornputer-assisted image analysis was used to quan- the area corresponding to these colours. BSE images magnified at x50 corresponding to field widths of 1.8 mm were used to quantify the total porosity associated with each CP type.

The image andyzing program thresholds these colours with a pixel intensity range and

assigns a colour to each range. Table 6.3 lists the relationship between pixel intensity

range and material type.

Table 6.3 Relationship between pixel intensity range and material type

r Material 1 Sigmascan Pro Pixel intensity 1 BSE colour

1 Void area 1 Black 1 Red 1 0-56 1

Calcium phosphate LRW resin

White Grav

colour assignment Blue Green

range 188-255 119-174

Diagrammatidy, the following represents this relationship,

Figure 6.13. Illustration of a typical image generated in Sigmascan Pro fiom a BSE image. Blue = calcium phosphate, green = LRW resin and red = voids, correspondhg to pore openings.

Figures 6.14- 6.18 shows the various images captured, fiom a typical sample of each CP

type analyzed, spliced together in Corel Draw 5.0 to create a montage. The montage

represents the suface area that was cornputer-analyzed. The infiltration procedure was

not completely successful in the Zimmer and CAM Implant samples as indicated by the

gray areas observed within the pore volume of marcopores located in the centrai region of

the sample. However, the pores located dong the perimeter of the Zirnmer and CAM

Implant samples were completed infiltrated with resin. Figure 6-18 shows the stnits that

comprise the interconnecting porous network of the CPP material. The CPP material

showed complete resin intiItration as denoted by the absences of gray areas.

Figure 6.14. Montage of as-ceceid Zimmer nirÉrc a f k r LRW infibation. Field width = 3.6 mm.

Figure 6.15. Montage of as rccievcd CAM40/60 sdàcc after LRW infiltration. Field widàth = 2.6 mm.

Figure 6.16. Montage of as received CAM70/30 suraice &et LRW infiltration. Field width = 4.2 mm.

Figure 6.17. Montage of as-received 2CAM70/30 surf'ace afker LRW infiltration. Field width = 3.0 mm.

Figure 6.18. Montage of as received CPP-2Oppi surface afler LRW infiltration. Field width = 3 Smrn

The total porosity was calculated by relatiag the areas corresponding to the CP

material, LRW resin and voids. In short, total porosity was measured by the following

Total Porosity = Void area x 100% CP material area + Void area + LRW resin area

The results are illustrated in Graph 6.6 with standard deviations represented by error bars.

zirnmer cam40/60 cam70/30 2cam70/30 CPP

Graph 6.6 Total porosity of as-received CP scaffolds. The porosity calculated includes both micro and macroporosity that were determïned using an image-analyzing program that quantified the void areas corresponding to the micro and macropores characteristic of each CP type. The bars represent standard deviation.

One-way ANOVA was used to determine statistical significance at p = 0.05 between the

total porosity calculated for the CP samples. Statistical si&nhficance in total porosity was

observed between the CPP-2Oppi material and the other calcium phosphate scafTolds (p =

0.000199). However, no statistical significance in total porosity between the Zirnmer and

CAM Implant samples was observed @ = 0.26). The CPP material displayed the highest

degree of total porosity among all the CPs compared. Table 6.4 lists the tabulated results

and ranking of the ceramics based on the extent of their total porosity. The highest and

lowest degree of porosity is assigned a score 1 and 5, respectively.

Table 6.4 Tabulated Total Porosity with Corresponding Rank

However, there was no statistical significance calculated between the Zimmer and CAM

Ceramic

ZUnmer C AM40/60 CAM70/3 O 2CAV70/30 CPP-20ppi

Implant samples in terms of total porosity, consequently, the TP rank wouid then be CPP

> CAM40/60 = CAM70/30 = 2CAM70/3O = Zimmer.

Total Porosity f lP ) 54 % 67% 61% 46% 85%

6.A.5. Degradation Behaviour of CP Scaffolds

Total Porosity (TP) b n k

4 2 3 5 1

6.A.S.a. Atomic Absorption Spectroscopy

The degradation profile of the CP samples was assessed by placing samples of

each type of CP supplied in O. 1 M Tris B d e r for 6 weeks. At each week, 4 samples were

rernoved fiom solution. One sample was prepared for SEM and the buffer solution fiom

three sample vials had their contents analyzed for ca2+ ions by atornic absorption

spectroscopy ( AAS). The ca2' concentration obtained fkom AAS was normalized with

the mass of its correspondhg CP sample. Graph 6.7 illustrates the ca2' concentration

leached out in to the buffer media over the 6-week study penod for the various CP

samples.

1 2 3 4 5 6 Time [w k]

Graph 3. Calcium [ca27 ions leached fiom CP samples incubated in 0.1 M Tris bufEer (pH 7.4) for 6 weeks. Atomic absorption spectroscopy was used to calculate the calcium ion concentration in the b a e r soluîion. The calcium ion concentration calculated was norrnaiized with the mass of the CP sample. The results show a variation in ca2+ leaching for the various CP samples over the 6-week study period The bars correspond to standard deviation.

One-way ANOVA was performed between samples and within sample types at each

week and over the entire study period Statistical significance was observed between

samples using one-way ANOVA. However, a t-test was perfonned to determine

statïstical sisnificance between all combinations involving two sample populations at

each tirne. Table 8 ranks the degradation behaviour (refîected by the amount of ca2+ ion

leaching) of each material compared with the other CP matexials at each week based on

Ta b k 6.5 Degradation Behaviour in O. 1 M Tris B a e r (pH 7.4)

3 1 Zimmer > CAM70/30 > 2CAM70130 > CAM40/60 > CPP 4 1 2CAM70/30 > CAM70/30 (= CAM40160) > Zimmer (= CAM40/60) > CPP (=

Weeks 1 2

Ranking Based on StatWtical Signiticance Zirnrner > CAM7OMO > CAM40/60 = 2CAM70/30 2 CPP * CAM70/30 = Zimrner > CAM40/60 = 2CAM70/30

There was statistical significance (p c 0.05) observeci in cafcium ion Ieaching over the

I

5 6

entire study period for CAM70/30, 2CAM70/30, Zimmer and CPP; however, no

CAM40/60) Zimmer > CAM40/60 = CAM70130 = 2CAM70/30 Zimmer > 2CAM70130 > CAM40160 > CAM70/30 > CPP

significance was observed for CAM40/60.

The symbol " 2 " is used to show that 2CAM70/30 = CPP but CAMW60 > CPP

The pH of the b e e r media was monitored over the 6-week study period Graph

6.8 shows the changed in pH of the 0.1M Tris Buffer solution as a bction of time for

the various CP materials.

Time [wk]

pZimmer

C A M40160

a CA M70130

p 2 C A M70/30

CPP-20ppi

Control

Graph 6.8 Change in 0.1 M Tris bufFer pH during the 6-week degradation study p e n d At each, the pH of the bufEer solution containhg a CP m p l e was measured using a pH meter. Standard deviations were calculateci and are represented by an error bar.

There was statistical significance (p c 0.05) observed over the entire study period for

each sample type- There was a &op in pH from the control pH at 1 week that was

statistically significant @ < 0.05). This was observed in ail CP samples. However,

observing the pH trends of the CAM Implant and Zimmer samples over the study period,

it appeared that there was a slow nse in pH up to 5 weeks, but this was followed by a

slight decrease at 6 weeks. The CPP sarnpîes showed a fluctuation on pH up to 4 weeks

but at 5 weeks there was a significant rise in pH that was maintained at 6 weeks-

6.A.S.b. SEM of Calcium Phosphate Surfaces during Degradation Shidy

In addition to caicuiating the calcium ion concentration in the b d e r media over

the entire study period, representative CP samples at 1, 4 and 6 weeks were prepared for

SEM. Figures 6.19 - 6.23 are micrographs that represent the various CP samples at these

time points. The micrographs of the Zimrner material reveaied similar surface

morphologies at 1, 4 and 6 weeks. However, there was considerably more debris

scattered on the surface as denoted by white particdate matter seen at 4 and 6 weeks than

at 1 week. Apparent at 1 week, but more clearly visualized at 6 weeks, was the presence

of large block-like structures embedded within the swface of the material that were

surrounded by smaller irregularly shaped grains. The micropores appeared to be larger

than those seen at 1 and 4 weeks, suggesting that the material was dissolving during the

study penod. The dissolution of the irregularly shaped CP grains allowed the exposure of

another CP grain structure comprising the Zimrner material. Based on SEM images

generated fiom the Zimmer material published in literature (Dziedic et al., 1996), the

smaller ùregularly shaped grains are indicative of the TCP phase and the larger block-

like siruchires depict the HA phase.

The individual grains that comprise the CAM40/60 were observed to have

changed morphologically over the time points studied (Figures 6.2OA - L). Observing the

surface of the sample at 1 week, the individual grains appeared spherical. They were no

longer smooth as they appeared prior to incubation (Figures 6.9A -D), but rather the

sphencal shape was maintained by the organization of thin leaflet structures (Figure

6.20C). At higher magnification (Figure 6.20D), there appeared to be a deposition of

irregularly shaped white structures between the leafïetç. The calcium phosphate grains

within the surface pore volume appeared more sphencal in shape and had a smoother

surface (Figure 6.20B). This smooth area was surrounded by thin leaflet structures

sirnilar to those comprising the surface grains. At 4 weeks, there was little evidence of

leaflet structures on the surface of the material or within the pore volume. The grains still

remained sphencal in shape (Figures 6.20E). However, at higher magnifïcaîion, there

appeared to be white particdate matter covering the surface of the individual grains

resulting in a rougher surface appearaace (Figure 6.20F and H). At 6 weeks, a

distribution in grain sizes was observed with the largest and smallest grains measuring

approxirnately 15 and 2 Pm, respectively (Figure 6.201). Again, the spherical grain

structure was maintained and there was little evidence of thin leafiet structures

surrounding single grains.

The morphology of the individual grains of the CAM70/3 0 samples observed at 1,

4 and 6 weeks had changed over the degradation study period. At 1 week, the grains were

sphencal and smooth but at 4 weeks the grains were no longer smooth (Figures 6.21B

and E). The exposed surface of the single grain cornprised thin leatlet structures that

appeared to extend out fiom the bulk of the grain. The sphencal morphology of the grains

was longer evident at 6 weeks (Figures 6.2 1G) but rather the grains were surrounded with

thin leaflet structures and had their surfaces scattered with white particdate matter

(Figures 6.21B and C).

The 2CAM70/30 grain structures showed complete morphologicai change during

the degradation study pend, as observed at 1 ,4 and 6 weeks (Figures 6-22A - I). The

entire surface of the material constituted thin leaflet structures that appeared to introduce

another level of microporosity than that observed in the as-received scaffolds (Figure

6.9M - P). At 1 and 4 weeks, the spherical shape of the grains was still apparent.

However, this shape was no longer seen at 6 weeks (Figures 6.22H). The surface of the

material had a perforated appearance at this t h e point. At higher magnïkïcation, there

appeared to be a deposition of small globular-like matter on the exposed surfaces in the

structures compnsing the material (Figure 6.221).

There was no apparent morphological change in the CPP grain structures at 1, 4

and 6 weeks (Figures 6.23A - 1). The grains appeared similar to those obsewed of the as-

received material throughout the study period.

Figures 6.19A - L. Scanning electron micrographs of Zimmer samples incubated in 0.1 M Tris b a e r . Micrographs A - Dy E - H, 1 - L show increasing maenification of the surface of the Zimmer samples exposed to the buî€er solution at 1, 4, and 6 weeks, respectively. The erosion of the surface due to degradation has aiiowed the exposure of larger grains seen at 1-week (B) but more clearly evident at 6 weeks (0. The precipitation of white particdate matter on the surface of the sample was evident by 4 weeks (G) but more diamatically seen at 6 weeks (I and J). It would appear that the srnalier grains cornprishg the material had dissolved as seen by the increase in microporosity nom 1 @) to 6 weeks &). . F.W. = 90 p, 18 pm, 9 pn and 9 pm for A, By C and D, respectively. F. W. = 90 pm, 18 pm, 9 pn and 9 p for Ey F, G and & respectively. F.W. = 18 pu, 18 p, 9 pn and 9 pm for 1, Jy K and L, respectivefy.

Figures 6.20A - L. Scanning elextron micrographs of CAM40/60 samples incubated in 0.1 M Tris buffer. Micrographs A - D, E - H, 1 - L show increasing magnincation of the surface of the CAM40/60 samples exposed to the buffet solution at 1, 4, and 6 weeks, respectively. The change in individual grain structure over 1, 4 and 6 weeks is clearly seen in micrographs C, G, K. The prestnce of microcrystals, seen as white particdate matter, was first seen at 1 week @). By 6 weeks, there was an increase in the prevalence of thin leaflet structures surrounding individual grains, as seen in L. F.W. = 90 Pm, 18 p, 22.5 p and 6 pm for A, By C and D, respectively. F.W. = 90 pm, 18 pm, 9 pn and 6 p for E, F, G and H, respectively. F. W. = 90 pn, 36 p, 18 Cm and 18 pm for 1, J, K and L, respectively.

Figures 6.21A - 1. Scsinning electron micrographs of CAM70/30 samples incubated in 0.1M Tris bufEer. Micrographs A - C, D - F, G - 1 show increasing magnification of the surface of the CAM70/30 sample exposed to the b a e r solution at 1, 4, and 6 weeks, respectively. The change in grain morphology was clearly seen over the study period, as denoted in micrographs C,F, and 1. The smooth rounded appearance, seen at 1 week, has become roughened by the appearance of thin leafkt structures situated verticaily on the surface of individual grains (F) and surrouuding the grains, seen at 6 weeks. F.W. = 90 p, 18 pm, and 9 pm for A, B, and C, respectively. F. W. = 90 p, 18 pm, and 9 pm D, E, and F, respectively. F.W. = 90 pn, 36 pm, and 18 pm for G, H and 1, respectively.

Figures 6.22A - 1. Scaoning electron micrographs of 2CAM7OB0 samples incubated in 0.1M Tris buffer. Micrographs A - C, D - F, G - 1 show increasing maguification of the surface of the 2CAM70/30 samples exposed to the b d e r solution at 1, 4, and 6 weeks, respectively. At 1 week, the individual grains comprising the 2CAM70/30 sample have completely undergone morphological change compared to the as-received grain structure (Figures 13M -P). The individual grains are seen to be fomed of staggered thin leaftlet plates (A-C) that dimish in thiclcness and prevelance by 6 weeks (H and Z). The appearance of white particdate matter within the microporosity of the matenal is clearly seen at 6 weeks o. F.W. = 90 p, 36 p, and 18 pm for A, B. and C, respectively. F.W. =90 p, 18 un, and9 pm D, E, andF, respectively. F.W. = 90 p, 18 p, and9 pm for G, H and 1, respectively.

Figures 6.23A - 1. Scanning electron micrograpbs of CPP-20ppi samples incubated in 0.1 M Tris bufKer. Mifrographs A - C, D - F, G - 1 show increasing rnagnification of the surface of the CPP-2Oppi samples exposed to the buffer solution at 1, 4, and 6 weeks, respectively. It is evident nom micrographs A, D, G, that there is no significant change in microporosity and individuai grain size, as seen in B, E and H. The white particdate matter observed on the d a c e of a grain, seen in E, are an aaifacts of critical point drying. F.W. = 90 pm, 18 pm, and 9 Fm for A, B, and C, respectively. F.W. = 90 p, 18 p, and 6 pm D, E, and F, respectively. F.W. = 90 p, 18 CM, and 6 pm for G, H and 1, respectively.

6.B. Static In Vitro Studies

6.B.1. CeU Culture Substrate for Cell Seeding and Colonization

It was necessary to determine an appropriate substrate onto which a 3-

dimensionai material could be placed in order to seed cells. The appropriate substrate

would be one that would maximize ceii colonization on the 3-D matenal. Two substrates

were used in this expriment: 1) tissue culture treated (TCT) and 2) bacteriological-grade

(BG) 24-we11 plates. Representative samples of each type of calcium phosphate supplied

(Zimmer, CAM40/60 and CPP-20ppi) were placed onto TCT and BG substrates and

seeded with BMDC for 2 weeks. BMDC were also seeded directly on to TCT and BG

substrates to serve as positive and negative control, respectively. After 2 weeks, TCT and

BG surfaces as well as the bottom face of the ceramic (the surface in direct contact with

the two substrates) were analyzed by SEM. The activity of the BMDC seeded duectly

ont0 K T and BG substrates is clearly seen in Figures 6.24A-B and 6.25A - B,

respectively. Figure 6.24A shows evidence of globular-like structures, indicative of

cernent line matrix, which appears to cover the surface of the TCT substrate. At higher

magnification (Figure 6.24B), collagen bundles were seen to lie over top this globular-

like surface and become encrusted with irregularly shaped crystal-like structures

indicative of calcium phosphate crystals. in contrast, Figure 6.25A shows the state of the

BMDC population grown on BG for 2 weeks. Irregularly shaped structures were seen

colonizing the BG surface. These structures are likely the remnants of dead cellular

matter that have failed to spread on this substrate.

The surface in direct contact with the TCT and BG substrates was analyzed to

assess the viability of the ceLi population grown at this interface. Figures 6.24C -F and

6.25C - F show the morphologieal appearance of the cells colonizing this sudace.

Figures 6.24A - F. Scanning electron micrographs of ce11 colonization and activity on 3- D porous cetarnics cultured in TCT 24-well plates. Passaged rat bone marrow cells were seeded on to 3-D porous substrates and cultured for 2 weeks in TCT 24-well plates. RBM cells were seeded directly on to TCT weUs to serve as positive controls. Figure A shows a ce11 layer pulled away after processing exposing globular-like accretions that appear to make direct contact with the TCT substrate. At higher magnifïcation, Figure B, mineralized collagen was seen encrusted into this globular-like layer. Figure C shows the 3-D porous surface that was in direct contact with TCT surface for 2 weeks. At this magnification, it appears that there is minimai cell colonization on this surface and at higher magnification, Figure D, the cells appear necrotic. Loose comective tissue that does nct appear mineralized was seen colonizing this surface. Figure E shows calcium phosphate particles, located away fiom the bulk sample. encrusted with mineralized collagen (Figure F). Such particles are LikeIy the result of the degrading calcium phosphate block sample. F.W- = 60 Fm, 18 p, 450 p, 60 Pm, 60 mm and 1 1.4 prn for A, B, C, D, E and F.

Figures 6.25A -F . Scanning electron micrographs of ce11 colonization and activity on 3- D porous ceramics cultured in BG 24-well plates. Figures A and B reveal the state of the ce11 population cultured directly on to the BG substrate at low and tugh magnifications, respectively. M e r 2 weeks, the cells appeared rounded and very few were seen spreading on the BG surface. Figure C shows the underside of a 3-D porous ceramic that was in contact with the BG surface for 2 weeks. Very few cells were observed on this surface and loose connective tissue was seen colonizing this surface. Figure D shows a calcium phosphate hgment that has become loose fiom the buik sample. There was no evidence of cellular colonizattion or collagen mineralization on or around this particle. Figure E illustrates the BG surface that was in contact with a CPP sample. Again, there was no evidence of cellular colonization or activity in this area. However, Figure F shows the surfàce of the CPP that was in direct contact with the BG surface during the 2-week culture period. There was evidence of cellular colonization and activity throughout the porous structure. Cellular matter appeared to be deposited in to the micropores of the sample with ce11 sheets over-laying the surface grains. F.W. = 180 Pm, 45 pm, 600 Pm, 150 Fm, 0.9 mmand45 pnfor A, B, C, D, EandF.

6.B.2. Optimum Ce11 Seeding T h e

The appropriate ce11 seeding time was detemiined by counthg the total number of

cells adhering to TCT nibstrate as a function of time. Graph 6.9 iilustrates the total

amount of adherent and non-adherent cells counted at 0.5, 1,3 and 4 hours.

& --- --- - ---- - ---------

0 -5 1 3 4 Time [hi]

O NAC

W'C --

Graph 6.9 Total celi attachent to TCT (tissue culture treated) plastic as a function of time. Passaged rat bone marrow cells were seeded on to TCT 24-well plates for 0.5, 1, 3 and 4 hours. At each time point, 0.01% trypsin was used to remove the adherent cells from the substrate surface. A Coulter counter was used to count the total ce11 number correspondhg to the adherent cells (AC) and non-adherent cells (NAC). The bars represent standard deviation.

It is evident with increasing t h e that the amount of adherent cells counted increased

while the non-adherent cells counted decreased. These results suggest that the minimal

acceptable ce11 seeding time on TCT is 1 hour. Consequently, seeding passaged RBMD

ceils for 1 hour was considered the optimum ce11 seeding time for funw experiments.

6.B3.CeU Adherence & Colonization on 3 4 CPs during CeU Seeding

To study the capacity of cells to adhere and colonize 3-D calcium phosphate

substrates during the 1 hour ce11 seeding period, the total number of cells adhering to the

CP surface, to the well-plate (bacteriological grade) and suspended in ce11 culture media

were counted. Three different CP substrates were used (Zimmer, CAM40/60 and CPP-

2Oppi) as well as negative (BG) and positive (TCT) control substrates. The results of the

cell attachment assay are presented in Graph 6.10.

- - - - --

- NAC

m*C Weil-plate

Graph 6.10 Total ceil attachent to various substrates after t hour. NAC, AC and Well- plate columns represent the total ce11 number counted via a Coulter counter in the cdhue media, attached to the substrate and well-plate, respectively, after 1 hour of ce11 seeding. TCT plastic and BG (bacteriological grade treated plastic) was used as positive and negative controls, respectively. The error bars correspond to the calculated standard deviation.

To assess the success of the trypsinization procedure, a sample of each CP type,

after trypsinization, was stained with toludine blue. Figures 6.26A-D are light

photographs showing the various surfaces o f the CP stained with TB. At higher

magnifications, obtained by SEM, the presence of cellular debris remaining after

trypsinization was clearly visualized on the various CP substrates (Figures 6.27A - F). These observations suggest that the trypsinization procedure used in the cell attachment

assay was not a reliable method of quantirjting ce11 attachment to CP substrates.

Figures 6.26A - D. Colour light photographs showing the various surf's of the CAM40/60 and Zimmer samples stauied with toluidine blue afler trypsinktion. Figures A and B show the surfaces of the CAM40/60 and Zimmer, respectively, that were seeded with cells for 1 hour, trypsinized for 45 minutes and then stained with toluidine blue. A blue-violet colour was seen on this ceii seeding surface, with an increase in colour intensity seen within the pore volume. Figure C shows the fracture surface of the CAM40/60 sample and Figure D shows the surface opposite the ce11 seeding surface of the Zimmer sample. There was no evidence of cellular penetration within the bulk of the CP samples or cell colonization of the surface opposite the seeding surface, as denoted by the lack of blue-violet staining seen in Figure D. Photographs A, B, C and D have field widths = 23 mm.

Figures 6.27A - F. Scaoniag electron micmgraphs showing the Zimmer, CPP and CAM40/60 surfaces pst-trypsinizttion. Figures A - B, C - D, and E - F show increasing magniûcations of the Zimmer, CPP and CAM40/60 d a c e s after trypsinktion, respectively. At low magnifïcation, there was evidence of cellular matter stiil present on the surfàce of the various CP substrates and at a higher magnïfication, whole cells were seen invaginating the microporosity of the surface 0. F.W. = 180 p, 36 pm, 600 p, 45 p, 180 pm and 22.5 pm for A, B, C, D, E and F.

To assess the extent of colonization on the 3-D CP substrates at 1 hour,

representative samples of each CP type were prepared for SEM. The morphological

appearance of the cells colonizing a representative CP substrate (CAM40/60) is seen in

Figures 6.28A - D. Most of the cells present on the various calcium phosphate d a c e s

appeared as round cells extending short microspikes and larnellipodia. The seeding

surface appeared to be compietely covered with these cells making it difficult to observe

the underlying substraa

Figures 6.28A - D. Scanning electron micrographs of cell colonhtion on 3-D porous ceramics after 1 hour of ceil seeding. Figures 34A - D are micrographs of increasing m a ~ c a t i o n s showing the morphology of the cells colonizing the surface of a CAM40/60 seeding s h c e . Simila. observations were seen on the Zimmer, CAM70/30, 2CAM70/30 and CPP surfaces. Field widths = 300 pn, 300 pm, 45 p and 18 p for A, B, C and D, respectively.

6-B.4- Dexamethasone (O) Culfures

Calcium phosphate samples mauitaiaed in the absence of DEX for 7 days did net

display signs of osteogenic activity in the fom of mineralized bone tissue. This

observation is consistent with previous reports on the rat bone marrow system, which

state the requirement of DEX in the culture media for osteoprogenitor dinerentiation and

matrix synthesis. After 1 week in DEX (-) culture conditions, the surfaces of each porous

CP type were investigated for signs of osteoclastic resportion. Evident by SEM

observation of the Zimmer material (Figure 6.29A - D) was the presence of large

aggregate cells with round morphologies. At higher ma@cation (Figure 6.29C),

irregular shallow erosions on the calcium phosphate surface in the vicinity of these cens

was suggestive of resorption. The size of the resorption lacunae was measured to be

about 10 - 30 Pm. Most pits were round having scalloped borders, but multilobular pits

were occasionally observed. The cells occupying these pits appeared to have a rough

surface morphology (Figure 6.29D) perhaps indicative of muitinuclearity and/or the

intemalization of calcium phosphate particles. It is inconclusive; however, whether these

cells are osteoclasts resorbing the calcium phosphate surface. Consequently, the

micrographs merely suggest that these cells are osteoclast-like cells. A positive TRAP

(Tartrate Resistant Acid Phosphatase) activity wouid aid in contiming the multinuclear

phenotype of osteoclasts but was not conducted as part of this project.

The CAM Implant and CPP ceramics did not show evidence of active resorption

of their surfaces by giant multinuclear cells a d o r osteoclast-like cells. Instead their

surfaces were covered with thin ce11 sheets that appeared fibroblastic in nature (Figures

6.29E, 1, M and R). These fibroblastic-like cells were spread out over the entire surface

area exposed and made contact with the calcium phosphate grains through inserthg their

pseudopodia înto the microporsity of the materid (Figures 6.29F, J, O and T).

Figures 639A - T. Scanning electron rnicrographs showing the celi population coloninng the surfaces of the CP sample types d e r 1-week incubation in DEX (-) culture conditions. Micrographs A- D, E - H, 1 - L, M - P and Q - T show increashg magnifications of the ceil population interacting with the Zimmer, CAM60/40, CAM70/30, 2CAM70/30 and CPP samples incubated in DEX(-) culture media Ceil morphology on each CP d a c e is observed in C. G, J, O and R. In 33C, celis possesshg d e d borders? indicative of osteoclast-like cells, are seen colonizing the Zimmer surface. However, in R, the ceils appear fibroblastic on the surface of the CPP sample. F. W. = 45 pn, 36 Pm, 18 pm and 9 pm for A - D, respectively. F. W. = 180 pm, 45 q, 36 p m and 18 Fm for E - H, respectively. F.W. = 180 Pm, 45 pm, 18 Pm and 18 pm for I - L, respectively. F-W. = 90 pm, 90 Fm, 45 pm and 18 pn for M - P, respectively. F-W. = 300 Pm, 180 Fm, 180 pm and 90 pm for Q - T, respectively.

6.C. CeU Colonization, Arrangement and Function in Static & Dynamic Culture Systems as a fùnction of Tirne

6.C.l. SEM Observations in the Static Culture Environment

The CP samples at 4 hours, 2 days and at 1 - 6 weeks were observed by SEM to

shidy ce11 colonization, arrangement and activity throughout the macroporosity

characteristic of each sample type. Observing the seeding surface of the Zimmer and

CAM Implant samples after 4 hours of incubation at 37"C, 95% air and 5% CO2, the

osteogenic cells with plump rounded morphology appeared to colonize the entire surface

area (Figure 6.30A - D). These cells invaded even shallow surface pores. At the edge of

the samples, the cells appeared to be more flat and migratory as hdicated by extending

pseudopodia. The migration of the cells appeared to be facilitated by the anchoring of

their pseudopodia to the substrats (Figure 6.30C). After 2 days, the osteogenic cells

assumed a more elongated, fibroblastic-like morphology (Figure 6.30K). At 2 days, the

fkst sign of ce11 bridging pores was observed. Figure 6.3 1A shows an extending ce1

process bridging an approximately 170 p m pore and Figure 6-3 1 C illustrates an entire

ce11 body bndging an approximately 100 prn pore opening with its cell processes making

contact with the periphery of the pore volume (Figure 6.3 ID). It was apparent at this

early stage in the ce11 culture that the osteogenic cell population had colonized the

calcium phosphate surface and had become migratory after 4 hours. However, there was

no evidence of early bone formation events (Le. deposition of cernent line by osteoblasts)

at these early t h e points. These observations were consistently seen on dl CP matenals.

Figures 6-A - L. Scanning eltctton micrographs of ce1 colonization, migration and activity on 3-D porous ceramics at 4 hours and 2 days. Passageci RBM celis were seeding on 3-D porous substrates for 4 h and 2 days in static cuItutiag environments. Figures A - D and 1 - L are micrograp6s of increasing maglllfication showing the morphology of ~e celi population at 4 h and 2 days, rrspectively. Figures E - H are micrographs of inmeashg maguifïcatio~ls illustrating the morphology of the ccIls colonizing a 3-D porous substrate a f k 4 houn of dynamic culturing. The morphology of the cells coloniang the same 3-D porous subsûate cultuceci in static and dynamic environments a . 4 hours appeared similar. After 4 hours the celis appcated migratory as indicaîed by their extcnding pseuQpodia (Figure B). A&er 2 days, the cclls have s p d over the entire sading surEace and have migrateci to the edge of the sample (Figure 1 and K). Extending micmspikes and lamciüpodias wexe observed at 2 days (Figure L). F.W. = 600 p. 45 p. 45 pm and 45 pn for A - D, rrspctively. F.W. = 600 pm, 300 pm, 180 pm and 60 fot E - H, rcspcctÏvely. F.W. = 450 CL^, 90 pm, 90 pm and 36 pm for I - L, respectiveiy.

Figures 631A - D. S d g elcctmn mictograpbs of c d bridging at 2 days. Figure A shows a ceii ptocess extending over a pore opehg of appmrimately 170 p after 2 days of ceii culture. At higher maoiiflcation, Figure C, it was evident that the ceii process was extending h m a singie ce11 body. Figures C and D show a whole c d extending over a 100 pm pore opening a f i 2 days at low a d higber magnifhtions, rcspectively. F. W. = 290 un, 1 10 pan, 290 pm and 58 pn for A - D, respectivee1y.

Examining the surface of the Zimmer and CAM Implant samples at I week, it

was apparent that the rnajonty of the surface pores became occluded with ceU bridging

the pore openings, as observed by SEM. The cells covering the pores had interacted to

form a thin sheet that appeared to spiral around the pore volume resulting in its complete

coverage, as seen in Figure 6.32C. After 6 weeks, the surface of the material had become

completely covered with multi-layered ce11 sheets (Figure 6.32D). Observing the fieeze-

fractured surface of the samples at 6 weeks (Figure 6.32E - H), it was apparent that the

ceil sheets surrounded the entire exposed surface area with no evidence of cellular

penetration and colonization within the bulk of the sarnple. At higher mapification

(Figures 6.32F), pore bndging of almost d l the surface pores by ce11 sheets was

visualized. However, there was evidence of loose comective tissue present within the

pore volume but bone tissue formation was not observed. These observations were

consistently observed in the Zimmer and CAM Implant ceramics at al1 the time points

studied. In contras& the CPP-45ppi, 20ppi and IOppi scafliolds, after 1 week in DEX (+)

culture, had ce11 layers covering their entire porous structure that followed the contour of

the pore openings with no evidence of pore bridging (Figures 6.3 3 A - C). Occluded pores

(Figure 6.33D), however, were observed in Mly intercomected CPP samples of a

smaller pore size range (1 50-200 pm)-

Figures 632A - L. Scaaoing electron micrographs of the colonization of cells on 3-D porous substrates after 6 weeks. Figures A - H are micrographs illustrating the colonization of celis on 3-D porous cetamics cultured in a static environment- Figure A shows the typical Zimmer a d o r CAM ImpIant porous surface that was seeded with RBM cells. After 1 week in DEX (+) culture conditions, ce11 sheets were seen overiaying these surface pores (Figure B) and at higher magnification, Figure C, it is apparent that ce11 sheets are spiraling over the pore volume resulting it its complete occlusion. At 6 weeks, the entire surface area was covered with multi-layered ce11 sheets (Figure D) and the surface pore openings were no longer visible. Figure E shows a fieeze-fractured surface of a typical CAM implant sample cultured with RBM cells d e r 6 weeks. At this magnification, it was evident that ce11 sheets bad fomed dong the surface area of the block sampte but not any evidence of cellular activity was seen within the bulk material. Evidence of pore bridging was seen over a pore opening measuring 150 pm (Figure F). However, some surface pores contained in their pore volume loose comective tissue that are overlaid with ce11 sheets (Figure G), but at higher magnification, there was no evidence of cement line formation or mineralized collagen (Figure H). M e r 6 weeks in the dynamic culture environment, ce11 sheets had also grown over surface pore openings and followed the contour of the block cerarnic sample (Figure 1). However, where intemal pores were seen connected to surface pores, there was evidence of loose connective tissue colonking the pore volume (Figure J and K). At higher rnagnincation, Figure L, it is apparent that the deposition of the loose connective tissue was made possible by the intercomecting channels between pores. F.W. = 2.25 mm, 2.25 mm, 693 Fm and 2.25 p m for A - D, respectively. F.W. = 2.58 mm, 0.9 mm, 528 p m and 60 p for E - H, respectively. F.W. = 360 pm, 1.00 mm, 1.30 mm and 450 pm for 1 - L, respectively.

Figures 6.33A - D. ScaMing electron micrographs of CPP samples cuitured in DEX (+) static culture conditions at 1 week. Figures A, B and C show the colonization of celis dong the entire porous network of the CPP-lûppi, CPP-2Oppi and CPP-45ppi samples, respectively. In Figure D, there îs evidence of cell bridging over fbily htercomected pore openings having nominal diameters measuring 150 p. F.W. = 2.32 mm, 1.74 mm, 1.74 mm and 2.8 mm for A - D, respectively.

On examining various portions of the fieeze-fiactured surfaces, it is evident that

d l calcium phosphate sarnples supported bone tissue formation. The evidence of cernent

line formation, coiiagen miwralization and morphologicaily distinguishable bone tissue

observed on the entire CP substrats demonstrates this. Figure 6.34A shows an elongated,

fibroblast-like osteoblast lying over mineraüzed collagen bundles on the Zimmer

subsate (also observed on the CAM Implant and CPP samples). While on the same

sarnple surface, an osteocyte is seen lying in its fonning lacuna (Figure 6.34B). The

centre of the fkeeze-hctured surface showed no evidence of biological matter (Figure

6.34C) rather the HA and TCP grallis that comprise the materiai were cleariy

distinguishable.

Figure 6.34A-C. Figure A shows a scanning electron micropph of an osteoblast migrating over minetalized coilagen colonizing the calcium phosphate substrats of the Zimmer material at 1 week (F.W. = 44 p). Figure B shows a scanning electron micrograph showing an osteocyte embedding itself in its newly forming l a c m after 2 weeks (F.W. = 12.8 p). Figure C shows a scannuig electron micrograph showing the central area of the fieeze-fracturai surface of the Zimmer material after 2 weeks in static culture conditions. At this magnincation, only the grains that comprise the materiai are seen with no evidence of biologicai matter in the vicinity (F. W. = 29.7 p).

The different CAM Implants ceramics studied showed similar biologicai profïie.

An amorphous layer, measuring approximately 0.5 pm, covered the exposed spherical

grains comprising the CAM Implant materials. This layer was deposited dong the

contour of the grain structures as well as on the thin leaflet structures surroundhg the

sphencal grains, as seen in Figures 6.35A and B. At 6 weeks, morphologically

distinguishable bone tissue was seen only on the exposed surface of the material, above

the pore volume and never seen within the pore volume or in the bulk of the material

(Figure 6.36A).

Figure 6.37 represents the appearance of the CPP scaEold after 2 weeks cuitwe in

the presence of RBMD cells. These cells populated the outer and inner surfaces of the 3-

D scaffold and formed a continuous sheet following the surface contour of the scaffold as

seen in Figure 6.37A. Figures 6.37B, C and D provide higher magnifications of scanning

electron micrographs of the matrix laid down within the scafTold. Figure 37B

demonstrates mineralized collagen fibres bridging between individual grains of the CPP

scaffold. At this magnification the collagen would seem to be inserting into the

microcrystalline surface layer of the CPP grain structure, which is more clearly visuaiized

in Figure 6.37C as a surface layer which is quite distinct fiom the underlying CPP

surface. Ce11 processes aiso adhered to this surface layer, as shown in Figure 6.37D,

which enveloped the grain structure of the underlying surface. These results were

consistent with al1 macroporous CPP types studied.

6.C.2. SEM Observations in the Dynamic Culture Environment

After 6 weeks in the dynamic cell culture system employed, the surface pore

volumes of fieeae-fiactured Zirnmer and CAM implant samples showed more evidence

of cellular matter and morphologicaily distiaguishable bone tissue (Figures 6.36C) that

than that observed in the static environment. Some surface pores were occluded with

overlying ce11 sheets, as observed in the static culture environment. However, the ianer

surface of the pores were lined with an afiibillar layer u p n whïch collagen bundies were

mineralized (Figures 6.35C). Densely packed mineralized tissue, characteristic of bone

tissue occupied the remainder of the pore volume. This new bone tissue appeared to make

intirnate contact with the calcium phosphate grains, as shown in Figure 6.35D.

Figures 6.35A- D. Cernent üne formation in static and dynamic culture enviroments. Figures A and D are scanning electron micrographs of fkeeze-fkactured surfaces of the CAM40/60 sample cultured in DEX(+) static and dynamic environments d e r 6 weeks, respectively. There was evidence, at this magnincation, of cernent h e formation that appeared to be deposited dong the surface of the grains. After 6 weeks of static culturing, an amorphous-lüre layer was seen covering the surface of the grain structures and had been deposited within the spaces between individual grains (Figure B). Figure D shows densely packed bone matrix that had formed d e r 6 weeks of dynamic culturing that appeared to make intimate contact with the grain structure. F.W. = 6.5 pm for A - D, respectively.

Figures 636A - D. Scannuig electron micrographs showing the extent of bone matrix elabmation on 3-D porous ceramics cuitured in static and dyoamic environments. Figures A and C depict keze-fkactured surfaces of 3-D porous ceraxnics showing morphological distinguishable bone grown in static and dynamic culture conditions, respectively. The bone ma& depicted in Figure A had grown over a surface pore opening while the bone forrned in Figure 42C had grown within îhe pore volume of a surface pore. Figures B and D show evidence of mineralized collagen colonizing the surface of the calcium phosphate grains. F.W. = 43 pn, 13 pu, 43 p and 20 pn for A - D, respectively.

Figure 6.37A - D. Scanning electron micrographs showing the appearance of the CPP scaffold cultuced in the presence of rat bone marrow celis for 2 weeks. F.W. = 178 ~ i m , 6 p, 2 pm, 3.4 pm for A, B, C and D, respectively-

6.C.3. LM of CeU Colonization and Arrangement in Static & Dynamk Cuhre Systems

Senal histological sections were cut according to the method described previously

of the CP porous samples at 1 - 6 weeks and stained with H&E. The samples were seeded

(approximately 1 x 106 total cells) with subcultured RBMD cells and grown D M (+)

culture both in static and dynamic media environments. Figures 6-38A -D are H&E

stained histological sections representing the midpoint (approximately 2 mm into the

sarnple) of ceII-seeded CPs cultured in static and dynamic systems at 6 weeks. The white

areas indicate the ghost of the ceramic after decalcification. However, even after 5 weeks

in decalcimg solution rernnants of calcium phosphate crystals were observed

(tramlucent particdate matter). In particufar, CAM70/3O and CPP sarnples took

approximately 8 weeks to decalcifjr, although some crystals still remained. It is clearly

evident that there was a greater abundance of bone-like tissue (dark pink) covering the

surface of samples cultured in the dynamic environment than in the static environment.

Figures 6.41B shows plump osteoblasts linuig the s d a c e of newly fonned bone that

produced substantially more matrix and bone than that observed on the CPs cultured

statically. There was considerably more mineralized tissue observed within the surface

pore volumes of the Z h e r and CAM Implant samples grown in the dynamic culture

system than in the static system (Figures 6.3912 and B). It is apparent in Figures 6.38C

and 6.38D that there is new bone formation in the intemal pores, htercomecting

channels and the surface of the CPP sarnple both in static and dynamic systems.

However, more bone formation was observed on CPP samples grown in the 3-D rotating

system. The Azan Heidenhain connective tissue stain was used to clearly confirm bone-

like tissue comprishg coilagen (dark blue). The nuclei of cells appeared red as a result of

this stain. Figure 6.42C shows the abundance of bone tissue with embedded osteocytes

laid down within the pore structures of the CPP ceramic cuitured dynamically.

As detennined previously, there is a varying degree of interconnectivity

associated with the CP porous ceramics suppiied. Observing the interior portion of the

Zimmer and CAM Implant samples, it was evident that there was loose comective tissue

occupying the pore volume as denoted by the light pink stain (Figures 6.40A). However,

no bone formation was observed witbin the bulk of these materids. RecaII that the

samples were cut to the midway point and the histological sections presented, are

therefore, representative of this portion of the sample. In cornparison, samples cultured

dynamically showed a greater degree of cellular matter and comective tissue within the

pore volume and, in some areas bone-like tissue were observed (Figures 6.40B).

Examining the cellular activity at 2, 4 and 6 weeks on representative samples of

the Zimmer, CAM Implant and CPP types cultured in static and dynamic environments, it

was evident that the cell-seeded scaffolds incubated in the d y n e system supported

more bone mat& formation than that observed in the static environment at al1 t h e

points. In fact, even after 8 weeks of static culturing, there appeared to be considerably

less matrix deposition dong the surface of the Zimmer and CAM Implant scaffolds than

that observed after 6 weeks of dynamic culturing (Figure 6.43).

Figure 638A-D. Six weeks d e r static and dynamic cell culturing of porous calcium phosphate substnites (HM stain). Figures A and B show the amount of bone matrix elaboratioa (red stain) on the surface of a CAM Implant sample cultured in çtatic and dynamic environments, respectively. Figures C and D show bone r n h elaboration throughout a porous CPP-20ppi sample maintained in static and dynamic culture conditions. White area indicates the ghost of the cerarnic remaining &er decalcification. Figures A, B, C, and D have field widtbs = 683 W.

Figures 639A - B. Extent of pore bridging on 3-D porous ceramic surfaces d e r six weeks of static and dynamic ceil culturing (HE sbin). Figure A (F.W. = 109 pm) shows thin cell layers bridging surface pores (indicaîed with an arrow) having small amounts of bone matrur within the pore volume (static culture), while Figure B (F.W. =

109 pm) shows a higher degree of bone matrix formation within the pore volume of surface pores (dynamic culture). (CAM Implant-type shown here)

Figures 6.40A - B. Lack of osteogenesis within the bulk of the CAM Implant samples (Azan Heidenbain connective tissue stain). Figures A (F.W.= 109 p) and B @.W.= 109 pm) show a histological section representïng the midpoint of the bufk sample cultureci in static and dynamic environments, respectively. Both figures dernonstrate the lack of osteogenesis throughout the entire porous network, as indicated by the lack of staining in the central portion of the sample. Similar observations were seen in the Zimmer samples. White area indicates the ghost of the ceramic d e r decalcification,

'(w =wl) auoq arlr am=!pam pu p e y w =nq3ws aql Aq (qtmpq sir 'aumlorr aiod a q q p \ pqmd Alaspap s? tirq) auoq atp JO a p s atp %uy! uaas 9s a = i q m o %wax ' ( ~ ~ o L z =-MX '3 ngwr sga mq,y pappaqm (UW pi) sa~A30;)3so a q pw ( q s aqq) anssp auoq paqpa- JO aawsad atp smn~rim u y s anssg a+vouum meqnap!aH mzv a u -aaoq pauuoj A ~ M ~ u ay) uqaw pppquua

a- s a W m '(d OLZ = -fi-d .Bis 3 ' 4 ~ ~0 (8) uoge3@em 3a@q iv ' ( sa ldm ~ammz tq riaas osp) saldmas x q d q yw3 JO aumlorr aiod a q u g p uog~mio~ auoq $0 ald->ta p!dh B MOqS ( d n 601 = 'M'J 3 3 H) V !S33ll%d * h ~ ( l @ ~ @ p p m p saidmes a JO samnlon aiod uopmuoj auoa -3 -VI*, sam2hg

3

Figure 6.42A - D. Conhation of bone formation by the Azan Heidenhain comective tissue s t a h Figures A @.W. = 683 pm) and B @.W. = 683 p) show typical examples of bone elaboration on CPP-20ppi and CAM Implant/Zimmer samples cultureci dynamicaily. Figure C (F.W. = 109 pm) reveals that histolgically identifiably bone has fomed throughout the porous CPP sample (blue stain). While in Figure B, bone is formed o d y dong the surface of the CAM Implant and sometimes within the pore volume (Figure D, F.W. = 109 p ). ( A h seen in the Zimmer samples).

Figures 6.43A- D. Extent of bone matrix elaboration on CP scafEolds after 8 weeks of static culturing (H&E). Figure A demonstrates that even after 8 weeks of static culturing there is still less rnatrix laid down on the surface of the CP scaffiolds (Zimmer-typ shown here) than after 6 weeks of dynamic cuitining (Figure B). It wodd appear that there is comparable amount of bone matrix deposition on the s d k e of the CP sdKolds after 4 weeks of dynamic culturing (Figure C) and 8 weeks of static culturing (Figure A). Figure D shows the extent of rnatrix deposition after 4 weeks of static culturing. Boue maaix is shown as a darkened band (indicated with an arrow in A) situateci dong the CF substrate. Fields width for A, B, and D = 0.27 mm.

6.C.4. BSE Imaging

Figure 6.44 shows a montage of backscattered scanning electron images of the

surface of a CAM40/60 sample &er 6 weeks in ceil culture. The white areas represent

the calcium phosphate components, the gray areas depict the embedding resin and the

black areas are the pores not infïltrated with resin. Light gray areas were observed around

the circumference of the sample and sparingly seen in some pore areas. Scannïng electron

rnicroscopy confirmed that these areas were biologicd. Pore bridging by ce11 sheets, also

observed by SEM and LM, were observed along s d a c e pores by BSE imaging. The

cellular behaviour of pore bndging occurred along d a c e pores in both Zimmer and

CAM Implant samples. It was observed that pore bridging occurred over pores of certain

d e r and outer diameters. Consequently, using 12 randomly chosen Zimmer and CAM

Implant samples cultured for 6 weeks with RBDC, the outer and inner diameters of

surface pores that were non-bridged and bridged were tabulated. Surface pore dimensions

were calculated fiom BSE images taken midway through the samples. The relationship

between the outer and inner diameter of a pore (Ofli) provided information about the

potential for that pore to become bndged. In order to interpret the results, two

assumptions were made: 1) ce11 bridging was independent of the CP composition and 2) t

15 % threshold range was valid. Calculating the Od/Oi relationship (Appendix D), the

following interpretations were made: 1) non-bridged pores were observed at = 1 zk

15 % or 0.85 - 1.15 threshold range for non-bridged pores if Od > 152 pm and/or Od > Oi,

2) bndged pores were observed at Od/Oi = 1 + 15% if Od < 229 pm ancilor Od < Oi and 3)

for al1 other data on pore dimensions, bridging was observed if Orni < 0.85 and >

1.15.

Figure 6.44. Mmtagc of BSE images of the CAM40/60 surfàce a f k 6 wceks ofall culture. The arrow incikates ce11 sheets, Field width = 4.2 mm-

6.D. In Vivo Studies

Retrieval of the CPP scaffiolds fiom rat femora after 2 weeks demonstrated that

bone had grown throughout the pore structure as evident in Figure 6.45. Figure 6.45A

provides an overview of the site of implantation with respect to the neighboring cortical

bone, although at this magnification, it is difficult to distinguish the implant fkom the

reparative trabecdar bone. Figure 6.45B iUustrates the distinction between bone

containing an osteocyte lacuna and the neighboring implant with its characteristic grain

structure. It should be noted with these fieeze-fractured samples that the spaces appearing

between the biological surface tissue and the CPP are artifacts of processing. Figure

6.4SC shows that a bone seam having grown dong a surface of the CPP scaffold is

anchored through the interdigitation of the bone matrix with surface rnicroporosity of the

scaiTold. It is aiso evident that some of the surface grains of the scaffold, which have not

been influenced by the fieeze-hcturing technique, exhibited a rounded rnorphology.

This morphological change was evident on both the lateral and apical facets of individual

grains. The ability of the elaborated biological matrix to envelope the CPP scaff+old is

s h o w in Figure 6.45D, where individual grains can be seen surrounded by a network of

collagen fibres.

Histological sections findings fkom the CPP implants retrieved &er 6 and 23

weeks indicated that bone formation occurred on the surface of pore regions and

advanced toward the centre of the pore. in addition, new bone formed in direct apposition

to the implant surface (Figure 6.46B). At 23 weeks, giant multinuclear cells were seen on

the implant surface (Figure 6.47A and B). However, evidence of active resorption was

mattainable.

Figure 6.45A - D. Scanning electron micrographs showing the CPP scaffolds in vivo retrieved f i e r 2 weeks. The implantation site is illustrateci in A. Field widih = 1.7 p. The appearance of the CPP grain structure and its relationship with the biological siaface is seen in B. (Note also the osteocyte at the bottom right). Field width = 43 p. The interdigitation of the bone matrix with the surface of the scaffold is evident in C. Field width = 32 p. Individual CPP grains enveloped with collagen fibres are seen in D. Field widîh = 18 p m.

Figura 6.46A - B. Twenty-three weeks d e r transfemoral implantation of CPP samples (H&E stain). Figure A (F. W. = 195 p) illustrates the implantation site (bottom-centre) and the newly fonned bone. Despite decalcification protocols, remnants of the CPP material are clearly seen after 23 weeks (Figure B. F.W. = 109 pm). These remnants represent the struts fonning the porous network Re-ts of the CPP material (white porous matter, labeled CPP) are seen intimately contacting the newly formed bone (marked with arrows). At this magni.&ation, no evidence of fibrous tissue is seen interfâcing the CPP material and the newly fomed bone.

A B

Figare 6.47A - B. Evidence of osteoclastic resorption of CPP scaffolds implanted in rat femora after 23 week ( H E stain). Figures A and B show gimt multi-nucleated ceUs sitthg on the surface of CPP remnants d e r decalcification (marked with amw). However, it is not evident in these ligbt rnicrographs thai the cells are actively resorbing the CPP material. Field widtbs for A and B = 270 pm.

7. DISCUSSION

The purpose of the work reported hzrein was to assess the enicacy of various

calcium phosphate scaffolds for tissue engineering. Specifically the hypothesis addresses

the need of interco~ecting macroporosity as an essential prerequisite for TE

applications. The results have unequivocally demonstrated that 3-D interconnecting

porosity is essential for the successful employment of a calcium phosphate TE scaffoId.

This discussion therefore focuses on the following major fïndings of the work reported:

scaf5old composition; degree of interconnecting macroporosity; scafEold degradation by

dissolution or cellular means; in vitro biological characterization in b o t . static and

dynamic culture systems; and the in vivo response of an ideal CP scafTofd.

7.A. Physical Characterization of Candidate TE Scaffolds

7.A.1. Scaffold Composition

Considerable effort has focused on investigating the suitability of using calcium

phosphates as synthetic bone grafts and more recently, as scafSolds for bone tissue

engineering applications (Ohgushi et al., 1989: Goshima et al., 1991% b; Yoshikawa et

al., 1996). Severai stoichiometries of calcium phosphates have been investigated for their

poteniial healing role in bone repair (de Groot et al., 1992; Toth et al., 1995; Tampien et

al, 1997). It is important, therefore, to determine the composition of the calcium

phosphate that is to be applied to bony sites since its composition is believed to influence

its biocompatibility, osteoconductive potential, mechanical properties and biodegradation

profile (Jarcho et al., 198 1; LeGeros et al., 1995; O'KelIy et al., 1996).

In the present work, powder x-ray diffraction spectroscopy was used to confim

the identity, crystallinity and phase purity of the calcium phosphates supplied. The XRD

spectra confïrmed that the Zimmer and CAM Implant samples used in this study were

biphasic calcium phosphates comprising 2-phase mixtures of HA and TCP. However, the

TCP phase of the CAM70130 sample contained both a and P - TCP polymorphs and the

minor phase of the 2CAM70/30 was confirmed to comprise the a-TCP polymorph.

Sintering between 900°C and 1 100°C has resulted in the formation of P-TCP with the HA

phase and higher sintering temperatures (>1300°C) has caused the formation of a-TCP

(LeGeros et al., 1995) In contrast, the XRD spectnim generated fiom the CPP sampie

revealed that it was pure P-CMP (calcium metaphosphate, Ca(P03)2 ). This composition

was expected since the starting material in the CPP production is non-crystalline CMP

that becomes crystalline after sintering at 900°C for 1 hout (Baksh et al., 1998). AU

samples generated XRD spectnuns that confumed their crystallinity-

7.A.2. Porosity & Interconnecting Macroporosity

Porosity that includes both micro and macroporosity is important to consider in a

scaffold design since it effects the degradation and biological properties of the materiai.

In the bioceramic processing field, microporosity relates to the spaces that are left when

the powder particies are not compIeteIy joined after sintering and macroporosity relates to

the larger pores that measure greater than a few microns (Klein et al., 1983).

Macroporosity has been shown to be important in providing rigid fixation of the implant

to the skeletal system by allowing bone ingrowth into the porous structure (Cameron et

al., 1976; Klawitter et al., 1976; Spector et al., 1976). These early reports using porous

cerarnic and polymer grafts suggest that the optimum rate of bone ingrowth was observed

in pore sizes approximately 100-135 pm (Klawitter et ai., 1976). For bone tissue

engineering applications, both the degradation profile and a matrix that permits bony

ingrowth in vivo needs also to be considered in the scaffold design. Particularly, the

material once grafted into the patient should degrade in a manner that is coincident with

bone remodeling; that is, the implant should impart mechanical stability during new bone

formation but, itself should degrade during remodeling.

Scaoning electron microscopy was used to show the micro/macroporosity and

Ievel of interconnectivity characteristic of the CP simples received. The micrographs

revealed that there was a wide range of rnacropore sizes found in the Zimmer and CAM

Implant samples with the smallest pores measuring 50 pm and the largest meamring

1200 Fm. The nominal pore size calcuiated for the Zimmer, CAM40/609 CAM70/30 and

2CAM70/30 was 400, 600, 325 and 300 pm, respectively. The pore diameters were

generally irregular in shape and very few were seen to connect to neighboring pores. The

highest degree of intercomectivity was seen in the CPP materials that displayed the

highest degree of intercomecting macroporosity. The nominal pore size calcuiated for the

CPP-45ppi, CPP-2Oppi and CPP-lOppi was 450, 850 and 1000 Fm, respectively.

Specifically, the resultant CPP45ppi and CPP-2Oppi macrostnicture appeared similar to

that observed of human trabecular bone and therefore, satisQ the criteria of mimicking

the replacement tissue.

It is apparent that the resultant architecture of the various samples is a reflection

of their processing route. Specifically, the macroporous CPP scafTolds were made using

polyurethane (PU) sponge method (Lee et al., 1996). The PU sponge was burnt out and

the resultant inorganic CPP matrix remained. The processing routes of both the Zimmer

and CAM Implant samples are not described herein due to confidentiality agreements.

However, it can be speculated that the pores of the ceramics were created by a procedure

similar to the Hubbard method (Toth et ai., 1995). This method involves mixing the

calcium phosphate powder with sked naphthalene beads. When the naphthalene

sublimes, pores are left behind that retain the size of the naphthalene beads. Another

possible method of pore creation relies on the decomposition of hydrogen peroxide to

generate a pore-nlled structure (Jarcho, 1981). Despite the processing route used, the CPs

supplied posses a pore size range that is suitable for bony ingrowth or osseointegration of

the implant; that is a minimum nominal pore size of 80 pm (Mainard et al, 1996).

However, in the in vitro milieu the pore size range for bony Uigmwth has been shown to

be different. Matrices with a nominai pore size of 200 pn resulted in occlusion of pores

by migrating cells (Rout et al., 1987). Consequently, in the in vitro stage of the TE

strategy, the rate and distribution of osteogenesis around and within the porous implant

will Vary considerably, depending upon the macropore size as well as the size and

number of intercomecting channels.

It is evident that al1 the CP ceramics produced have undergone a second stage

hvolving thermal treatment since the resultant grains appeared to be sintered together.

The sintering process has bonded the particles together, which is only achieved at high

temperatures. Sintering parameters for calcium phosphates Vary from temperatures of

1 O00 - 1 300°C and times of 1 - 24 hours (Toth et al., 1995). The microporosity evident

fkom the high magnification micrographs taken of the various CP sarnples, measuring a

few microns, has been created due to the gaps left between the sintered particles. This

microporosity created is usefûi for creating strong bone-implant mechanical interlockhg

interactions as a consequence of bone growth into these micropores (Dziedic et al.,

1996).

n i e conventional method of m e r c q porosimetry was not used to calculate totai

porosity since this method is restricted to pore sizes measuring 75 p (Schugens et al.,

1996). Light photography confïrmed that the CP scaEolds supplied had average pore

sizes greater than 75 p. Consequently, cornputer assisted image analysis was used to

quantity total porosity. It is evident fiom the results (Graph 2) that the CPP-20ppi

scaffolds had the highest degree of total porosity when compared to the Zirnmer and

CAM Implant scafEioIds. However, the totd porosities measined tiom Zimmer and CAM

Implant scaffolds proved not to be statistical signifïcant.

The totai porosity (micro and macroporosity) was calculated using an image-

analyzing program as described in the Resuits section. A similar approach was used to

quanti@ the interconnecting macroporosity. The following relationship was proposed to

represent interconnectivity:

interconnectivity = LRW resin area x 100% CP material area + Void area + LRW resin area

However, the results using the Zimrner and CAM Implant samples had no valuable

meaning in terms of interconnectivity since these materials displayed very few

interconnecting pores, as evident in Figures 6.14 - 6.17, which proved inconsistent with

the numbers generated by the above relationship. The montages created clearly show that

the materials were poorly infiltrated with resin which is a direct consequence of the lack

of interco~ecting pores that would otherwise result in successful infiltration as seen in

the CPP-20ppi montage. Interconnecting macroporosity becomes an important parameter

to measure for tissue engineering scaffold design since the maximum ce11 coverage

throughout the scafSold during the in virro stage will be governed by the extent of

interconnecting macrochannels. The ability of osteogenic cells to colonize the entire

surface area of the scaffiold will consequently influence its success as a TE constmct

when grafted into the patient. The methodology descnbed above proved hadequate to

measure intercomectivity since the image anaiysis software was incapable of

deconvoluting overlaying threshold ranges.

7.A.3. Degradation Behaviour

7.A.3.a Soiution-meàiated Processes

Any practical TE application involving the CP ceramics will involve contact with

a physiological environment, therefore, it is important to study the stability or

degradation of candidate materials. A material c a . degrade by solution-mediated

processes and/or by cell-mediated processes. Ln the present study both forms of

degradation were investigated. A O.1M Tris buffer solution was used to determine the

dissolution profile of the CP materials and a DEX(-) ceil culture environment was used to

determine the osteoclastic resorptive potentiai of the CP materials. Tris b a e r contains no

calcium and phosphate and therefore, any calcium measured during the study period was

soIely attributed to the calcium leached from the sample. During the 6-week study period,

al1 the CPs showed varying degrees of ca2+ leaching as analyzed by atomic absorption

spectroscopy. There was a fluctuation in ca2+ leachhg fiom the Zimmer material during

the 6-week study period. The CAM70130 and 2CAM70/30 materials showed a slow rise

in the level of ca2+ up to 3 and 4 weeks, respectively, followed by a significant drop in

calcium concentration at 5 weeks. The decrease in calcium concentration at the various

tirne points suggests that the dissolution of the ceramic caused a supersaturation of ca2'

and PO4 " ions in the microenvironment around the sample, resulting in precipitation of a

new calcium phosphate phase. The presence of this new phase perhaps caused the

increase in pH seen during the study period (Rohanizadeh et al., 1998). Scanning electron

micrographs generated at 1, 4 and 6 weeks for the Zimmer and CAM70130 samples

showed the precipitation of irregularly shaped microcrystals within the microporosity of

the ceramic surface perhaps of carbonated apatitic origin. In addition, there appeared to

be a signifïcant increase in microporosity and a decrease in average grain size observed in

the Zimmer material over the study p e n d Onïy &er 6 weeks were precipitated

microcrystals O bserved on the surface microporosity of the 2CAM7013 0 matenal. The

identity of the microcrystals was not detemiùied. However, it c m be speculated that these

microcrystais are CO3-apatitic in nature since carbonated apatite is known to fomi on the

surfaces of HA and TCP both in v i m and in vivo (LeGeros et al., 1995). The presence

of these microcrystals relates to the material's bioactivity and its ability to fonn a strong

bone-material interface (LeGeros et al., 1995). The nature of the bonding is likely due to

micro-mechanical interdigitation with the microtopography of the carbonated growth

phase (Davies, 1998). The SCAM70130 grains have undergone complete morphological

change during the 6-week study period. The change in morphology is likely attributed to

the presence of the a-TCP phase that has a higher dissolution rate than B-TCP (LeGeros

et al., 1995). The change in grain morphology combined with the decrease in calcium

concentration after 3 weeks suggests that calcium ions had reprecipitated on the surface

resulting in the growth in individual calcium phosphate grains, as denoted by the apparent

thinning and lengthening of the leaflet grains. The CPP and CAM40160 material showed

no trends in the amount of calcium leaching over the study period. Particdarly, the

rnicrographs generated fiom the CPP material incubated at 1, 4 and 6 weeks showed no

morphological change in grain structure or microporosity. The CPP material is made up a

polymenc structure consisting of phosphate chains that bind calcium ions through ionic

bonds. The method of degradation reported of CPP fibres Nt vin0 is by hydrolytic

degradation of the phosphate bonds via chah scission (Filiaggi et al., 1998). However,

the macroporous CPP scaiTolds used in this study did not show evidence of dissolution

during the 6-week incubation period in the buffer solution. Although, studies conducted

by Guo et al. (1994) using CPP fibres, repoaed that there was a reduction in CPP fibre

diameter with immersion time in Tris-buffered solution. These results suggest the

possibility of designing a macroporous CPP scaffold that can degrade in a tirne-

dependent manner that would render the scaffold suitable for the TE strategy. At first, the

scafTold wouid show rilinimai degradation so osteogenic cells could have a stable surface

to colonize. M e r implantation, the cell-seeded scaffold would continue to provide a

stable surface in order to Mpart mechanical stability, promote vascular invasion and

M e r ce11 colonization. At a certain defined tirne, the materiai would then begin to

degrade at a rate that coincides with bone remodeling until it is totally resorbed and

replaced with new bone tissue.

The scanning electron micrographs of the CAM40f60 revealed the presence of

precipitated microcrystais within the rnicroporosity of the material as early as 1 week.

Since the major phase present in the CAM40160 is P-TCP, it is likely that the greatest

amount of degradation occurred during the first week of the study period and that the

surpersaturation of ca2+ around the sample caused the precipitation of apatitic

microcrystals that are seen at 1 week. The results fiom this study demonstrated that in

vitro dissolution of biphasic ceramics was influence by the M C P ratios and the CPP

material showed no signiticant evidence of dissolution. However, the possibility of

programming the degradation rate into the CPP material design is plausible.

7.A.4. b. CeU-mediated Procases

The cell-mediated degradation of the CP supplied by osteoclasts was studied

using the DEX(-) rat bone marrow ceU culture system. Osteoclast cell types appear to be

the major cellular component involved in bone remodehg and hence, a material that can

be degraded by the action of osteoclasts would be ideal for bone TE applications.

Osteoclastic ce11 cultures are technically dficult since these primary cells do not

replicate and only remain alive for a very few days (Rey, 1998). Consequently, their

colonization and at tachent to the CP ceramics was evduated after 1 week in DEX(-)

ce11 culture conditions. Scannîng electron microscopy was used to show the

morphologicai and fûnctional characteristics of osteoclasts; that is, their multinuclearity

and d e d border formation on the surface of the calcium phosphate ceramic. The

micrographs generated showed evidence of aggregating giant cells on the surfaces of the

Zimmer and CAM Implant materials. In parcicular, the surface of the Zirnmer material

showed a greater nurnber of giant cells adhering to the surface that appeared to be located

in resorption pits. The underlying surface appeared to be degraded perhaps due to the

release of protons by the resportive organs of these cells. The morphology of the ce11

population appeared to be different on ail the CP surfaces. Specifically, the morphology

of the ce11 population appeared as giant muiti-lobular cells on the Zimmer surface, and as

flat fibroblastic-like cells on the CPP surface. The dBerent cell morphologies observed

on the various CP substrata suggest that the activity-state of this ce11 population is

effected by surface topography; this is a known phenomena described in some of our

previous work (Gomi, et alJ993). For resorption to occur, the osteoclast requires a fkm

attachment of the cell membrane to the substratum to effectively isolate the resorption

lacuna fiom the surroundhg media and permit the maintenance of the cell-generated pH

gradient (Lakkakorpi et al., 1991). Consequently, the heterogeneous composition of the

CAM Implant surfaces and their dissolution may have diminished the ability of these

cells to attach to, and then resorb, the surface. In addition, osteoclasts are motile ceUs

and the specidized nrfned border and sealing zone appear o d y when an osteoclast is

sessile and actively resorbing a surface (Lakkakorpi et al., 199 1). Consequently, perhaps

the cells colonizing the CAM Implant and CPP scafEolds cultured in DEX (-) conditions,

are merely migrating osteoclasts that are not actively resorbing the underlying

substratum.

7.B. In V i o Biological Characterization of Candidate TE Scaffolds

7.B.1. Cell Culture Techniques Involving Porous 3-D Substrates

Traditional ce11 culture techniques involve seeding passaged cells on to 2-

dimensional tissue culture treated surfaces in order to study the colonization, attachent

and activity of a single expanded celi population. However, the bone tissue engineering

strategy typically involves the use of 3-dimensional porous scaffolds that m u t be seeded

exclusively with the desired cell population in the in vitro environment prior to

implantation. Consequently, a suitable technique for seeding passaged cells on to 3-D

substrates in vitro must be employed. Particularly, cell-seeding and seeding tirne need to

be optimized for 3-D substrates.

On day 5 of the primary ce11 culture, rat bone marrow cultures were subcultured

ushg a trypsinization protocol and the maximum ce11 seeding density attainable from two

rat femora was prepared, this usually redted in a total cell count in the order of 106

(Appendix B). The device used to dispense the ceils on to the 3-D porous substrates was

a 10 ml syrïnge h e d with a 20G1112 precision glide needle. This gauge size was chosen

since this diameter opening is suitable for the passage of whole cells, averaging in 10 pm,

through the tip opening without rupturing ceil membranes. The technique used to seed

cells involved positioning the needle tip perpendicular to the substrate d a c e at the

centre of the sarnple. A constant pressure was applied to the syringe to dispense the cells

ont0 the surface. Employing this technique ensured that the ceils contacted the CP

substrate first pnor to contacthg the well-plate surface. To ensure that the cells adhered

preferentially to the CP substrate bactenological grade (BG) well-plates were used during

ce11 seeding. Bactenological grade well-plates have a hydrophobie ntrfaces that do not

promote ce11 attachent. Anchorage-dependent cells, such as osteogenic celis, require

hydrophilic surfaces in order to spread and migrate (Zygourakis, 1996). Consequently,

without adhering to a surface, differentiating osteogenic cells c m o t migrate and perform

their function. They take on a rounded morphology indicative of theu inability to spread.

In addition, bacteriological grade 15 ml round bottom tubes were used during the ce11

culture study. Round bottom tubes were chosen to permit ceil migration around the entire

sample surface and not restricting ce11 colonization to those surface exposed to the media

The resuit of culturing cells on 3-D porous substrates on both flat tissue culture treated

and bacteriological grade well-plates revealed poorly colonized cells on the surface

contacting the well-plate. In fact, the cells that colonized this surface appeared unhealthy

and unable to perform their fiinction; that is, secrete collagen and mineralize it even after

2 weeks. In healthy bone marrow ce11 cultures, mineralized collagen is seen as early as 1

week.

7.B.2. Appropriate Ceii-Seeding Time

It is important to define a suitable time to allow cells to adhere to a d a c e before

transferring the ceil-substrate complex to a permanent environment for long-tem

culturing. A 1 hour ce11 seeding time was chosen as the suitable t h e for permîtting ceH

adherence to a 3-D porous substrate based on the ce11 attachment assay utilizuig tissue

culture treated weli-plates (Ciraph 4). Ce11 seeding was performed by the technique

described above in bacteriological grade 24 well-plates. m e r 1 hour the ceil-substrate

complex was transferred to 15 ml round bottom mbes that were filled with 10 ml of fully

supplemented media.

Graph 5 shows the results of the total ce11 attachment to the various CP substrates

after 1 hour. A trypsinization protocol was used to detach the cells from the various

surfaces. It is evident that there were more total celis counted in the media (NAC) and

attached to the well plate than there were counted attached to the substrate (AC). The low

ce11 count attained fiom the substrate surface may initially suggest that the ceii seeding

technique employed was not suitable, but fiuther anaiysis revealed that the porous nature

of the scaffolds caused the ce11 suspension to flow out of the sample during ceil seeding

and tyrpinization was shown to be inefficient in removing al1 the cells fiom the substrate,

as seen in Figures 6.27A - F. Remnants of cellular debris and whole cells are seen

invaginating the microporosity of the substrate surface even after 45 minutes of

trypsinization. Consequently, the adherence of the cells resisted the action of the trypsin.

Despite the low number of cells attached to the surface, as counted by a CouIter counter,

the ce11 seeding technique employed was successfid in promoting ce11 colonization on the

CP substrates (Figure 6.28) and ce11 adherence to these substrates was stronger than the

action of trypsin.

7.B.3. CeU Colonization and Arrangement on 3-D scaffolds f i e r 4 hours of Static & Dynamic Culturing

DEerent 3-D calcium phosphates, havuig different porosities and

interconnectivities, were used as rat bone marrow ce11 culture substrates in order to

detennine whether the distribution of bone matrix formation in vitro would be a product

of their macrostructure- Bone tissue engineering in vitro involves the interaction of

osteogenic cells with a material surface. The nature of the substrate c m directly influence

cellular response, ultimately affecthg the rate and quality of new tissue formation

throughout the macrostructure. After 4 hours of ceU culture, osteogenic cells were seen

to colonize the ce11 seeding surfaces of the ceramics- The celis appeared migratory as

indicative of their extendhg pseudopodia. Apparently, the 3-dllnensional fluid flow did

not effect the migration of cells on the various CP surfaces, as indicated by the similarity

in ce11 morphology to that observed in the static culturing conditions at 4 hours (Figure

6.30A - H). The rough microtopography of the various CP surfaces may have faciliated

the formation of focal attachments (although not seen in SEM) between the cells and the

underlying substrate, permitting them to span the spaces between the grains and migrate

over the surface. The 3-D flow of media produced by the angle and rotation of the tube

around the central shaft did not result in ce11 detachment, as supported by the colonization

and migration of cells observed on the CP surfaces after 4 hours of dynamic culturing

(Figures 6.3 0E - H). Consequently, the cell-substrate attachent once established proved

to be stronger than the abiiity of fluid flow to detach these cells fiom the surface. This

the surface. This suggestç that the 1-hour ce11 seeding tirne performed in the static

environment provided the celis with the opportunity to establish strong interactions with

their underlying substrates. The focal adhesions made by the cells with their substrate

detennined their ceii shape that, when transduced via the cytoskeleton to the nucleus,

resulted in the expression of the specific ce11 phenotype seen in Figures 6.3OA - D

(Boyan et al., 1996)-

7.B.4. Pore Bridging & Occlusion

As dserentiating ostegenic cells migrate over a surface they are not onIy afEected

by the surface roughness and chemistry (Petite et al., 1996; Davies et al., 1997) but also

by the gross morphology that includes pore openings. At 2 days, there is evidence of ceU

prccesses extending over pore openings of 170 pm (Figure 6.3 1A and B) and ce11

bndging over pore size openings measuring 100 prn (Figure 6.3 1C and D). Osteoblasts

have been shown to prefer pore sizes ranging f5om 200 to 400 pn in diameter for

encouraging migration, attachment and proliferation in to the pore volume (Boyan et al.,

1996). This may be because the curvature of these pores provide optimum compression

and tension on the cell's mechanoreceptors that allow them to migrate into such pore size

openings. This occurrence of pore bridging seen at 2 days resulted in compete pore

occlusion by 1 week. Ceil-ce11 interactions fonned a thin sheet that spanned the surface

pore openings at 1 week and thickened by 6 weeks due to ce11 sheet multi-layering. Loose

connective tissue was seen colonizing the pore volume as demonstrated by light

microscopy. However, there was no evidence of bone matcix production withïn the pore

volume. For bone tissue engineering applications, the complete occlusion of the surface

pores poses a serious problem since pore occlusion prevents m e r cellular penetration

throughout the macrostructure that is necessary once the cell-seeded scaffold is implanted

in the patient. in addition, physiological fluid penetration, capillary invasion and

stabilization of the implant with bone ingrowth are prevented. The Zimmer and CAM

implant porous ceramics dernonstrated this ce11 phenomenon of pore bridging over pore

size openings measuring approximately 200 p m or less. However, this was clearly not

observed for the CPP-45ppi, CPP-2Oppi and CPP-lOppi scafTolds. Evidently, the CPP

materiais, having a high degree of uiterconnectïng macroporosity, possess the macropore

size ranges that are suitable for cellular migration and colonization of their entire surface

area. This suggests that Mly interconnected pores of the appropriate size will encourage

ce11 migration throughout the macrostructure. Of the CPP pore size ranges studied, it

would appear that a scaffold having a nominal pore size of 450 p m displaying fd l

intercomectivity would be sufficient in promoting maximum cell-surface coverage and

consequently, 3-D bone matrix production, as seen in Figure 6.33C, since the CPP-6Oppi

material (having nominal a pore size of 150 pm) despite its fiilly intercomecting

porosity, showed evidence of pore occlusion after 1 week in ce11 culture. Bone matrix

elaboration occurred on the surface of the CPP-6Oppi scaffold but no evidence of cellular

penetration was observed within the buik of the material. This observation was consistent

with the Zimmer and CAM Implant scafFoIds.

7.B.S. Osteogenic Activity on 3-D scaffolds maintained in Static & Dynamic Culture Systems

The CPP substrates d l supported in vitro bone formation as indicated by

morpho logicall y and histologicall y identifiable bone. However, the formation of bone

rnatrix was restricted to the six faces of the porous Zimmer and CAM Implant blocks.

The lack of bone matrix production within the interior of these substrates supports the

importance of a M y interconnecting rnacroporosity. The CPP substrates, excluding the

CPP-6Oppi substrate, supported 3-D bone ma& formation throughout the entire p rous

structure.

7.B.6. The Suitability of Dynamic Culturing

As present experience shows, the in viho stage of the bone TE strategy involves

three principal steps. The fïrst step of the strategy involves the cultivation and expansion

of bone marrow derived cells that is accomplished with the use of conventional tissue

cultured treated dishes. The second step entails seeding the expanded osteogenic ce11

population on to a suitable scaEold. A suitable technique for ceU seeding has been

proposed in this study (discussed previously) as well as an appropriate scafXold

pennitting 3-D ce11 spreading, namely the CPP scaffolds, has been established. The third

step aims towards the longterm maintenance of the differentiated osteogenic phenotype

on the 3-D scaffold. In the static environment of traditional culture dishes it is dïfflcult to

sustain longterm experiments without contamination and promoting differentiation and

maintenance of the desired ce11 population (Sittinger et al., 1997). Consequently, these

Iimitations prompted the development of a dynamic ce11 culture system that resembles as

close as possible the in vivo situation. The scaffolds were placed in tri-directionally

rotating ce11 culture media. Each sample was stabilized between stainless steel mesh

wires inserted at the centre of 50 ml conical tubes. In this position, the sample was

suspended in ce11 culture media and remained in complete contact with the culture media

throughout the culture. This configuration was used to 1) provide continuous distribution

of culture media throughout the porous structure and 2) to eliminate contact with the

walls of the tube since it has been reported that collisions with hard surfaces (Le. a wall)

may contribute to poor cell attachent (Qui et al., 1998). Histological fïndings

demonstrated the success of using the proposed dynamic culture system in sustaining a

viable ce11 population that encouraged more bone formation than that produced in a static

culture environment. Al1 CP porous samples demonstrated a greater quantity of

morphologically and histologically distinguishable bone matrix in the rotating media

environment. in fact, a greater nurnber of the surface pore volumes of the Zimmer and

CAM Implant samples were fiiled with bone matrix than wailed-off with ceii sheets, as

preferentially seen to occur in the static milieu. This suggests that the rotating fluid

environment has encouraged the migration of differentiating osteogenic cells within the

pore volume and the increased ce11 density within the pore has resulted in more ma&

production. However, there was stiii no bone matrix seen within the intenor of the

Zimmer and CAM Implant samples despite the rotating culture conditions. This again

supports the importance of a fully intercomecting scaffold for cellular penetration within

the bulk material that appears to be independent of culturing conditions. Once the cells

can be seeded, or are able to migrate, within the intenor of the scaSold, cuituring

conditions play a significant role in maintaining the ce11 population at the interstices. This

is supported by the r e d t s obtained by culturing osteogenic cells on the CPP scaffolds in

both static and dynamic culture conditions. The cells were seeded on to the scaffolds in

the same manner. However, after 6 weeks of static and dynamic culturing, the amount of

bone fonned throughout the macrostructure differed. In both enviroaments, osteogenic

cells colonized the entire surface area: however, more bone formed in the dynamic

culture. The rotating culture system provided continuous exchange of nutrients and

metabolic products fiom the scaf5old's microenvkonment to the sumounding fluid

environment. Whereas, in the -tic environment, there was perhaps a build up of

metabolic waste that remained stagnant around the scafTold's microenvironment that

consequently compromised the maintenance of the differentiated celI population and

hence, fûrther cellular activity. It has ken well established that osteoblasts respond to

hydrostatic pressure by altering their intemal structure that results in the upregulation of

bone matrix (Wilkes et al., 1996; Yoshikawa et al., 1997; Ingber et al., 1989).

Consequently, the physical stimulus of 3-D media flow thughout the scaffold has likety

translated into metabolic alterations that has promoted the differentiation of osteogenic

cells and enhanced bone formation. The dynamic ce11 culture system employed has

shown to sustah a more viable ce11 population at all the t h e points studied when

compared to the static ce11 cultures. Specifically, even after 8 weeks of static ceil

culturing, the amount of bone matrix formed on the various scaffulds was significantly

less than that observed after 6 weeks in the dynarnic environment. For tissue engineering

applications involving long term cultures on 3-D scaffolds, a dynarnic ce11 culture system

similar to the one proposed wouid promote greater ce11 stimulation that would result in a

high degree of ce11 differentiation. The cell-seeded scaffold wouid be grafted into the

patient to accelerate bone M i n g , therefore, a high degree of differentiated cells

colonizùig the entire scaKold would be pivotal in repairing a massive bony defect.

7.C. In Vivo bone growth throughout 3-D Scaffold

Macroporous scafTolds once grafted into a patient shouid not only be

biocompatible but also demonstrate the ability to support bone growth throughout its

porous network and fom an intimate bone-biomaterial interface. Extensive in vivo

studies have k e n conducted using the Zimmer material as both a graft substitute and TE

scaffold (Le. BMD cell-seeded scaffold) (Goshima et al., 1991% b and c; Ohgushi et al.,

1989; Kadiyala et al., 1997). However, despite its purpose (gr& or TE scaffold),

histolgicaüy findings have shown that only surface pores are filled with bone while pores

located in the centrai region are devoid of bone. Both the cells seeded on to the scaffold

in vitro and those present in vivo were unable to penetrate toward the buik of the scaffold,

as supported by histological findings (Ohgushi et al., 1992% b). Evidently, the lack of

bone uig~owth ia to these pores is a consequence of them king sequestered h i d e the

ceramic. Retrieval of the CPP scaffolds fiom the rat femora demonstrated that bone had

grown into the intergranular surface microporosity (Figures 6.45A- D) and also

throughout the porous structure, as seen histologicaily in Figures 6.46A - B.

Although the TE strategy was not performed in its entirety using the CPP scaffold,

the in vitro and in vivo observations suggest that the CPP scaffold is a suitable TE

candidate since it supports osteogenic ce11 colonization, migration and hc t ion

throughout its porous network in both environments. Consequently, M e r studies

involving the cell-seeded CPP scaf5olds impianted into bony defect sites need to be

performed to confirm the suitability of the CPP scaEold as a TE construct.

8. RELATING RESULTS BACK TO THE ORIGINAL HYPOl'HESIS

Issue 1: The extent of micro/macroporosity and interconnecting macroporosity

characteristic of each CP type studied will be a result of the processing procedure.

Validated: Scanning electron microscop y revealed the extent of micro/macropore size

range characteristic of each CP type studied. Although the processing procedure of the

Zimmer and CAM Implant scaf5old types are not reported herein, it is suggested that the

methods used to produce the porous scafEolds were different than that used to prepare the

CPP scaffolds. The CPP scaffolds were prepared using the PU (poiyurethane) sponge

method that resulted in various levels of interconnecting macroporosity that were not

observed in both the Zimrner and CAM Implant samples.

issue 2: The proposed method of quantification of interconnecting macroporosity by

cornputer-assisted image analysis wiil be effected by the extent of interconnecting

macropores present in the porous CP scaffold.

Validated for certain cases: Embedding the Zimmer and CAM Implant scaffolds in

LRW resin revealed the lack of fdly interconnecting macropores present in these sample

types. However, the CPP sample types were completely infiltrated with resin. The

proposed methodology using image analysis software to rneasure interconnectivity

proved hadequate for samples having poorly interconnecting macropores since the

software was incapable of deconvoluting overlay ing threshold ranges correspondhg to

fully, partially and non-infïltrated pore volumes.

Issue 3: Ce11 colonization on porous CP scaffolds will be affected by pore diameter by

resulting in cell bridging or ce11 migration into certain pore volume diameters.

Validated: Culhiring RBMD cells on porous CP scaf5olds resdted in pore bndging and

occlusion for scaffolds having pore sizes <229 p m and/or Od < Oi. Pore bridging was

seen after 2 days of ce11 culture resulting in complete pore occlusion observed &er 1

week. SEM and LM revealed very Iittle biological matter within the pore volumes of

surface pores of the Zimmer and CAM Implant scaffolds. The CPP scaffolds investigated

did not show signs of pore bridging at al1 t h e points studied.

Issue 4: The distribution of bone rnatrix formation will be a product of the macrostnicture

of the CP scaffold.

Validated: Ali CP scaEolds studied supported bone ma& formation. However, the

distribution of ma& formed on and throughout the porous network was inûueoced by

macroporsity and the extent of macmpore interconnections. RBMD cells colonized the

porous sdaces of the Zimmer and CAM Implant sample blocks, but due to the lack of

interconnections between the surface pores with the bulk pores, cells did not colonize the

buk sample. Ce11 migration into surface pore volumes were observed for pore diameters

> 152 prn andor Od > Oie

Issue 5: Dynamic cell culturing of RBMD cells on 3-D CP. porous scaffolds will result in

a greater degree of bone matrix elaboration when compared to culturing the same ce11

population statically on porous scaBolds.

Validated: Light microscopy revealed that there was a greater degree of matrix

elaboration on the Zimmer and CAM Implant scaffolds, and throughout the CPP

scaffolds, when RBMD cells were cultured on these substrates in the dynamic

environment system studied.

149

9. CONCLUSIONS

Ail the calcium phosphates supplied supported bone growth in vitra. Only the CPP

scaffold was employed for additionai experiments and was, again, shown to support

bone growth.

The highest degree of intercomecting macroporosity was found in the CPP scaffold-

types. The Zimmer and CAM Implant scaffolcis demonstrated a comparable level of

total porosity, but the macorpores present throughout the scaffolds were not fully

interconnected.

The distribution of the bone fomed throughout the scaffolds was a product of their

macrostructure. The CPP scaffolds demonstrated bone growth throughout their entire

porous network, while bone growth on the Zimmer and CAM Implant scaEolds was

restricted to the outer surfaces of the samples.

The proposed rotating culture method employed to create a 3-D fluid flow

environment enhanced bone matnx elaboration on al1 the calcium phosphate scaffolds

provided.

Appendix A

G.I. Composition & Preparation of FuMy Supplemented Medium

A.1.a. DEX(+) culture medium

a-MEM (Minimal Essentiai Medium)

15% FBS (Fetal Bovine Serurn)

10% Antibiotic : Penicilh G 167 units/ml Gentamicin 50 Arnphotericin B 0.3 p g / d

1 % Supplements: Dexamethasone 1 O-* M B-Glycerophosphate 5 mM L-Ascorbic Acid 50 pg/ml

Al1 the culture reagents above are mixed together to prepared fiilly supplernented medium (FSM). For 100 ml of F S M preparation, 75 ml of a-MEM was used.

A. 1. b. DEX(-) culture medium

Al1 the culture reagents used for DEX(-) culture are the same used for DEX(+) culture except dexamethasone is ornitted,

A.2. Composition & Preparation of O.lM Tris Buffer

O. 1 M m&C(CH20H)3] Tris ~ydroxymethy1)methylamine 6.0507g 0.01% waN3] Sodium azide 0.0500g

The above components are mixed together in 500 ml dm20 to prepare a stock solution of O. 1 M Tris buffer. The pH is adjusted to 7.4 using 1N HCI.

Appendix B

B.1. Calculation of Maximum CeU Seediag Density

Sample calculation:

Coulter counter readhg is made 3 times using one sample via1 containing 0.5 ml of passaged RBM cells and averaged.

Coulter counter reading averaged = 1700

Nurnber of ce11 in dilution (0.5 ml) = 1700/0.5 ml - - 3400 celldml

Number of cells in suspension = 3400 cells/ml x 20dOSml - - 1.36 x Io5 cells/ml

(Note: 20 ml is the volume of the suspension and 0Sml is the volume taken fiom the stock volume for counting)

Total number of ceils - - 1.36 x los cellsM x 10 ml - - 1.36 x 106

(Note: 10 ml is the total stock volume containing cells)

1-36 x 106 represents the maximum number of cells that are anainable h m subculturing cells harvested fiom 2 rat femora. The cell seeding density is obtained by dividing the total volume containing the passaged cells Le. 1 -36 x 1 o6 cell i lOml= 1 -36 x 10' ce l ldd

C.1. Preparation of Karnovsky's Fixative

For 25 ml Karnovsky's Fixative:

Dissolve 0.5 g parafoddehyde powder in 10 ml double distilled water, pre-heat to 56'C. Stir no less than 20 minutes. Maintain less than 60°C.

Add 1 drop of 1N NaOH and keep stir untiI solution is clear. Cool under tap water.

Add 2.5 ml 25% Glutaradehyde, 0.85 g Sucrose and 12.5 ml 0.2 M Sodium Cacodylate buffer. Adjust the W volume to 25 ml distiiled water.

Filter with #l filter paper and adjust pH 7.2 - 7.4.

D.1. Long-acting Ascorbic Acid

L-ascorbic acid 2-phosphate is used as a substitute for L-ascorbic acid in culture, which has the benefit of long term stability in aqueous solution. Equal concentration is used in BMC culture as routine L-ascorbic acid.

For l OOx L-ascorbic acid (5mg/ml):

MW = 176.1 dm01 Molar concentration = 5/ 176.1 = 0.0284 M

For making 1 OOx L-ascorbic acid 2-phosphate (0.0284 M):

Actual MW = MW 256.1 (fiee acid) + 1.5 mol mg + 4 mol H20 = 36455g

Note: Ce11 culture medium is supplemented with 0.000284 M of L-ascorbic acid 2- phosphate.

10. REFERENCES

Aubin, J.E., F. Liu. (1996) The Osteoblast Lineage. In Prhciples of Bone Biology, edited by J.P. Bilezikian, L.G. Raisz and G.A.Rodan. Academic Press, pp. 5 1-69.

Baksh, D., Davies, J.E., and Kim, S. Three dimensional matrices of calcium pol yphosphates support bone growth in vitro and in vivo. J Mater Science: Mater in Mëdicine 1 998;9:743-748.

Baron, R.E. (1 996) Anatomv and Ultrastructure of Bone. In Primer on the Metabolic Bone Diseases and Disorders of Mineral Metabolism, edited by M.J. Favus. Lippincott-Raven Publishers, pp. 3-38.

Barrows, T.H. Degradable implant matenals: A review of synthetic absorbable polymers and their application. Clin Materials 1986; 1 ~233-257.

Bauer T. W., Gessink, R.G.T., Zimmerman, R., and McMahon, J.T. Hydroxyapatite- coated femoral stems. Histological anaiysis of components retrieved at autopsy. J Bone Joint Surg 1991;73-A(10): 1439-1452.

Benahrned, M., Blottiere, H., Praloran, V., and Dacuisi, G. Monocyte activity in the presence of calcium phosphate activated by 1,25 (OH)z VD3 and interferon - 6. Biomaterials 1994; 1 5:25-30,

Berry, J.L., Geiger, J.M., Moran, J.M., Skraba, J.S., and Greenwald, A S Use of tricalcium phosphate or electncal stimulation to enhance the bone-porous implant interface. J Biomed Mater Res 1 986;20:65-77.

Boyan, B.D., Humrnert, T. W., Dean, D.D., and Schwartz, 2. Ro le of materiai surfaces in regulating bone and cartilage ce11 response. Biomaterials 1 996; 1 7: 1 3 7- 146.

Brekke, J.H., Toth, J.M., Vogelin, E., and Jones, N.F. (1998) Princi~les of tissue engineering governing osteoeenic devices. In Bioceramics Volume 1 1 (Proceedings of the 1 1" International Symposium on Ceramics in Medicine, New York, NY), edited by R.Z. LeGeros and J.P. LeGeros. World Scientific Publishing Co Pte. Ltd. pp. 341 -344.

Brown K.L. and Cruess RL. Bone and cartilage transplantation in orthopedic surgery: a review. J Bone Joint Surg [Am] l982;64:270-279.

Brown, P.W. and Constantz, B. Hydroxyapatite and Related Materials. Boca Raton, CRC Press, 1994.

Bmder, S. P., Fink, D. J., and Caplan, A.I.. Mesenchymal Stem Ceiis in Bone Development. Bone Repair, and Skeletal Regeneration Therapy. J Cell Biochem l994;56:283-294.

Cameron, H.U., Pïiiiar, RM. and Macnab, J. The rate of bone growth into porous metal. J Biomed Mater Res 1 976; 1 O:295-302.

Caneron, H.U. Evaluation of a Biodegradable Ceramic. J Biomed Maier Res 1977; 1 : 179-1 86.

Caplan, A.I. Mesenchymal stem cells. J Orthop Res 199 1;9:64l-650.

Caplan, A.I. and Bruder, S-P. (1 997) Ce11 and molecular engineering of bone re~eneration. In Principle of Tissue Engineering, edited by R Lanza, R-Langer and W. Chick. R.G- Landes Company, pp. 603 - 6 17.

Cook, S.D., Baffes, G.C. Wolfe, M.W., Sampath, T.K, and Rueger, D.C. Recombinant human bone morphogenetic protein-7 induces heaiing in a canine longe-bone segmenta1 defect model. Clin Orthop 1994;3O 1 :302-3 12.

Cook, S.D., Baffes, G.C. Wolfe, M N . , Sampath, T.K., Rueger, D.C., and Whitecloud, T.S. Effect of recombinant human osteogenic protein-1 on healing of large segemental bone defects. J Bone Joint Surg 1 997; 76A: 827-83 8.

Costantino, P.D., Friedman, C.D., and Jones, K. Experimental hydroxyapatite cernent cranioplasty. Plast Recomtr Surg 1992;90: 174- 185.

Daculsi, G. LeGeros, R.Z-$ Heughebaert, J-C., and Barbieux, 1. Formation of carbonate- apatite crystais after implantation of calcium phoshpate ceramics. Calcif Tissue Int 1 WOa;46:20-2%

Daculsi, G., Passuti, N., Martin, S., Deudon, C., Legeros, R-Z., and Raher, S. Macroporous calcium phophate cerarnic for long bone surgery in humans and dogs. Clinical and histogologicai study. J Biomed Mater Res 1990b;24:379-396.

Daculsi, G. and Passuti, N- Effect of the macroporosity for osseous substitution of calcium phosphate cerarnics. Biornaterials 1990c; 1 1 :86-87.

Davies LW., Tarrant S.F., and Matsuda T. (1987). Interaction between ~rimarv bone ce11 cultures and biomaterids. PART 1 : Method: the in vitro and in vivo stages. In Biomaterials and Clhical Applications , edited by A. Pizzoferrato P.G. Ravaglioli, and A.J.C. Lee, Elsevier, Amsterdam, pp.579-584.

Davies, J.E., Nagai, N., and Talceshita, N. (1989) Osteoaenesis and osteoclasis at the interfiace with a bioactive bone-substitute. In Trans. 15' Am. Meeting of the Society for Biomaterials, Lake Buena Vista, FI, USA, p. 5.

Davies, J.E., Chemecky, R., Lowenberg, B., and Shiga, A. Deposition and resorption of

calcified matrix in vitro by rat marrow cells. 199 1 a Cells & m e r 1 ; 1 :3- 15.

Davies, J.E., Nagai, N., Takeshita, N., and Smith, D.C. (1991b) Dewsition of cernent- linemaîrk on implant - materials. In The bone biomaterïal interface, edited by J.E. Davies, Toronto: University of Toronto Press, pp.285-294.

Davies, J.E. and Brady, J. (1992) A method to assess the reswnse of a bone substitute material to osteoclasis. CRC Handbook of Bioactive Ceramics, Vol. II, pp.283- 293.

Davies, J.E., Shapiro, G., and Lowenberg, B.F. Osteoclastic Resorption of Calcium Phosphate Ceramic Thin Films. Cells and Materials 1 993 ;3 :245-256.

Davies LE. In Vitro Modeling of the Bonelunplant Interface. A m t Rec 1996;245:426- 445.

Davies, J.E. and Baldan, N. Scanniug electron microscopy of the bone-bioactive implant interface. J Biomed MaterRes 1 997;3 6:429-440.

Davies, J.E. Mechanisms of Endosseous htegration. Inter J Prostho 1998; 1 l(5):391-409

de Bniijn, J.D., Klein, C.P.A.T.,de Groot, K., and van Blitterswijk, C.A. The Bone- Calcium Phophate Interface In Vitro. Biomaterials-Tissue Interfaces 1992; 1 O:305-309.

de Bruijn, J.D. (1993) Calcium Phosphate Biornaterials: Bone-bonding and Biodegradation Properties, PhD Thesis, Den Haag: Cip-Data Koninkilcjke Bibliotheek.

de Bruijn, J.D., van Blittenwijk, CA., and Davies J.E. Initial bone matrix formation at the hydroxapatite interface in vivo. J Biomed Mater Res 1995;29:89-99.

de Groot K. (198 1) Degradable ceramics. in Biocompatibility of Implant Materials, Vol 1, edited by D.F.Williams, CRC Press, Boca Raton, FL, USA, p. 199.

de Groot, K. Bioceramics of Calcium-Phosphate, CRC Press, Boca Raton, FL, 1983.

de Groot K, Geesink, R, Klein, C.P.A.T., and Serekian, P. Plasma sprayed coatings of hydroxlapatite. J Biomed Mater Res l987;2 1 : 13 75- 13 8 1 .

de Groot, K. (1988) Effect of mrositv and ~hysiochernical ~ ro~er t i e s on the stabilitv, resomtion, and strenmh of calcium phosphate ceramics. In Bioceramics: Material Characteristics versus In-Vivo Behavioru, AM. New York Acad. Sci., Vol. 523, pp.227.

de Groot, K., Klein, C.P.A.T., Wolke, J.G.C., and de Blieck-Hogervorst, J.M.A. (1 992)

Chemistrv of Calcium Phomhate Bioceramics. In CRC Handbook of Bioactive Ceramics Vol.II, edited by T.Yamamuro, LL Hench and J. Wilson. Boca Raton, Florida: CRC Press, hc., pp. 3-16.

de Lange, G.L., De Putter, C.,Burger, E.H., and De Groot, K. (1987) The bone- hvdroxvla~atite interface. In Biomaterials and Clinical Applications, edited by A. Pizzoferrato, P.G. Ravaglioli, and A.J.C. Lee, Elsevier, Amsterdam, pp. 217-222.

Denissen, H.W., de Groot, K., Ch Makkes, P., van den Hooff, A., and Klopper, P.J. Tissue response to dense apatite implants in rats. JBiomed Mater. Res 1980; 14:7 13-845.

Dick, H.M., Malinin T.I., and Mnaymneth W.A. Massive allograft implantation following radical resection of high-grade tumors requiring adjuvant chernotherapy treatment. Clin Orthop 1 985; 1 97: 88-95.

Dziedic, D.M., Sawa, I . H . , W i o n , D.S., and Davies, J.E. (1996) Osteoconduction on, and Bonding to. Calcium Phomhate Ceramic Imvlants. In Tram Mat.Res.Soc. Symp.Proc.Vol.4 14. Materials Research Society, pp. 147- 156.

Eschenroeder, H.C., McLaughlin, R.E., and Reger, S.I. Enchanced stabilization of porous-coated metal implants with tricalcium phosphate granules. Clin Orthop l987;2 16:234-246.

Filliaggi, M.J., Wells, J.D., Peel, S.A.F., Pilliar, R.M., Grynpas, M.D., and Kandel, R-A. (1998) Condensed calcium ~hosphate for soft tissue and bone repair/re~eneration. In BioceraxnÏcs Volume 1 I (Proceedings of the 1 l~ International Symposium on Ceramics in Medicine, New York, NY), edited by R.Z. LeGeros and J.P. LeGeros. World Scientific Publishing Co Pte. Ltd., pp. 34 1-344.

Flatley, T.J., Lynch, K.L., and Benson, M. Tissue response to implants of biphasic calcium phosphate ceramic in the rabbit spine. Clin Orthop R d Res 1983; 179:246-249.

Frayssinet, P., Hardy, D., and Rouquet, N. New observations on middle term hydroxyapatite-coated titanium alloy hip prostheses. Biomaterials 1 992; 13 :668- 696.

Frayssinet, P., Trouillet, N. Rouquet, Azimus, E., and Autofage, A. Osseointegration of macroporous calcium phosphate ceramics having a different chernical compostion. Biomaterials 1993 : l4:423-429.

Friedlaender G.E. Immune responses to osteochondral allografts. Clin Orthop 1985;197:115-121.

Fujirnori, Y., Mishima, K., Sugaya, T., Sakae, T., LeGeros, R.Z., Kozacoa, Y., and

Nagura, H. In vitro interactions of osteoclast-like celis and hydroxyapatite ceramics. In Bioceramics Volume II, edited by R.Z. LeGeros, Proceedings of the 1 1" International Symposium, New York, NY, USA, 1998, pp.335-338.

Gao T. J., Lindholm, T.S., Kommonen, B., Ragni, P., Paronzini, A., Lindholm, T.C ., Jamsa, T., and Jalovaara, B. Enhanced healing of segmental tibiai defects in sheep by a composite bone substitute composed of tricalcium phosphate cylinder, bone morphogenetic protein, and type IV coilagen. J Biomed Mater Res l996:32:505-5 12.

Geesink, R.G.T., de Groot, K., and Klein, C. P. A. Bonding of Bone to Apatite-Coated Implants. JBone andJoint Surgery 1988;70-B:17-21.

Geesink, R.G.T.Hydroxyapaptite-coated total hip prostheses. C h Orthop 1990;261:39- 57.

Goldberg, V.M. (1 992) Namal history of a u t o m and alloP;rafts. In Bone Implant Grafkg, edited by J Older. Springer-Verlag Berlin Heidelberg. p. 9-1 2.

Goldberg, V.M. and Stevenson, S. The biology of bone gr&. Sem in Arthrophty l993;43:58-63.

Gorni, K., Lowenburg, B., Shapiro. G., and Davies, J.E. Resorption of sintered synthetic hydroxyapatite by osteocIasts in vitro. Biomaterials 1 993 ; l4(2):9 1 -96.

Goshima, J., Goldberg, V.M., and Caplan, A.I. The osteogenic potential of cultured- expanded rate marrow mesenchymal cells assayed in vivo in calcium phosphate ceramic bloc ks. Clin Urthop Re1 Res 1 99 1 a;262:298-3 1 0.

Goshima, J., Goldberg, V.M., and Caplan, A.I. The origin of bone formed in composite grafts of porous calcium phosphate ceramic loaded with m m w cells. Clin Orthop Rel Res 199 1 b;269:274-3 1 1.

Goshima, J., Goldberg, V.M., and Caplan, A.I. Osteogenic potential of culture-expanded rat marrow cells as assayed in vivo with porous calcium phosphate ceramic. Biomater 199 1 c; l2;253-258.

Gundle, R., Joyna, C.J., and Triffïtt, J.T. Human bone tissue formation in d i f i i o n chamber culture in vivo by bone derived cells and marrow stroma1 fibroblastic cells. Bone 1995; l6(6):59%6O 1.

Guo, W., Kim, S., Grynpas, M.D., Prtitzker, K.P., and Pilliar, R.M. Calcium polyphosphates fibres for composite biomaterials - degradation study. In Tram. 2 0 ~ Am. Meethg of the Society for Biomaterials, Boston, MA, USA, p. 163.

Gross A.E., Lavoie M.V., McDemott P., and Marks P. The use of allogmft bone in

revision of total hip arthroplasty. Clin Orthop 1985; 197: 1 1 5- 121.

Gruber, H.E., Marshall, G.J., Kirchen, M.E., Kang, J., and Massry, S.G. improvements in dehydration and cernent Iine staining for methacrylate embedded human bone biopsies, Stain Technolagy l985;60:337-334.

Habal, M.B. (1992) Dflerent Forms of Bone Gr&. In Bone Transplantation edited by M. Aebi and Regazzoni. Springer-Verlag Berlin Heidelberg. pp. 6-8.

Hatterdey, G. and Chambers, T.J. Calcitonin receptors as markers for osteoclastic differentiation: correlation between generation of bone-resorptive ceils and cells that express receptors in mouse marrow cultures. Endocrinol 1989; 125: 1606- 1612-

Heiple K.F., Kendrick R.E., Hemdon CH., and Chase S. W. A critical evaluation of processed calf bone. J Bone Joint Surg [Am] 1967;49: 1 1 19- 1 127.

Hollinger, J.O., Brekke, J., Gruskin, E., and Lee, D. Role of bone substituts- Clin Orthop Rel Res 1996;324:55-65.

Holmes, R.E. Bone regenerations within a coralline hydroxyapaptite implant P l m Reconstr Surg I979;63 :626-633.

Holmes R.E., Wardrop, R.W., and Wolford, L.M. Hydroxylapatite as a bone graft substitute in orthognathic surgery: histologie and histometric hdings. L W Mc~xiIZofac. Surg. 1988;46:661-671.

Hoogendoorn H.A.,Renooij, W., Akkermans, L.M.A.,Visser, W., and Wittebol, P. Long- term Study of Large Ceramic Implants (Porous Hydroxyapatite) in Dog Femora. Clin Orthop Rel Res 1984; l87:S8 1-288.

Horowitz, M C and Fnedlaender. G.E. Immunologie aspects of bone transplantation: a rationale for fiiture studies. Orthop C h North Am 1987; t 8:227-247.

Holy, C.E., Shoichet, M.S., Campbell, A.A., Song, L., and Davies, LE. (1 998) Prevaration of a Novel Calcium Phosphate-Coated ScafTold for Bone Tissue Engineering. In Bioceramics Volume II, edited by R.Z. LeGeros, Proceedings of the 11' International Symposium, New York, NY, USA, pp.509-512.

Hughes C.W. Rate of absorption and callus stimulating properties of cow horn, ivory and beef bone and autogenous bone. Surg Gynecol Obsret 1943;76:665-671.

Hurley L.A., Zeier, F.G., and Stinchfield, F.E. Anorganic bone-grafting: clinicai experiences with heterogmfts processed by ethylenediamine extraction. Am J Surg 1960; 100: 12-2 1.

Ingber, D.E. and Follcmans, J. Mechanochernical switching between growth and differentiation during fibroblast growth factor-stimulated angiogenesis in vitro: role of extracellular matrix J Ce22 Biology 1989; 109(1):3 17-330.

Ishaug-Riley, S.L., Crane, G.M. Gurlek, A., Miller, M.J., Yasko, A.W., Yaszemski, M.J., and Mikos, A.G. Ectopic bone formation by marrow stroma1 osteoblast transplantion using poly@L-lactic-CO-glycolic acid) foams implanted into the rat mesentery. J Biomed Mater Res l997;36: 1-8.

Jaiswal N., Haynesworth, S.E., Caplau, A.I., and Bruder, S.P.Osteogenic Dflereniation of Purified Culture-Expanded Human Mesenchymal Stem Cells In Vitro. J Cell Biochem 1997;64:295-3 12.

Jansen J.A., de Wijn J.R., Wolters-Lutgerhorst, J.M.L , and van Mullem, P.F. Ultrastructural study of epithelial cells attachment to implant materials. Journal of Dental Research 1 985;64:89 1-896.

Jarcho, M. Calcium Phophate Ceramics as Hard Tissue Prothetics. Clin Orthop Re1 Res 198 1 ; 1 57:259-278.

Jarcho M., Kay, J.F., Kenneth, I,,Gumaer,K.I., Doremus, R.H., and Drobeck, H.P. Tissue cellular and subcellular events at a bone-ceramic hyhxylapatite interface. J. Bioeng 1977; 1 :79-92

Johnson, E.E., Urist, M.R and Fineman, G.A.M. Repair of segmental defects of the tibia with cancellous bone fi& augmentaed with human bone morphogenetic protein. Clin. Orthop 1988;236:249-257.

Jones, S.J., Boyde, A., and Ali, N.N. The resorption of biological and non-biological substrates by cultured avian and mammdian osteoclasts. Anat Embryol 1984; 170:247-256.

Katthagen, B.D. (1 986) Bone Regeneration with Bone Substitutes. Springer-Verlag, Berlin Heidelberg, pp. 1 -47.

Kadiyala, S ., Jaiswal, N. and Bruder, S .P. Culture-Expanded, Bone Marrow-Denved Mesenchymal Stem Cells Can Regenerate a Critical-Sized Segmental Bone Defect. Tissue Engineering 1997;3(2): 173- 185.

Kay, J.F. Bioactive Surface Coatings for Hard Tissue Biomaterials. in CRC Handbook of Bioactive Ceramics Vol.11, edited by T-Yamamuro, LL Hench and J-Wilson- Boca Raton, Florida: CRC Press, ïnc, pp. 1 1 1 - 122.

Kiawitter, J.J., Bagwell., J.G., and Weinstein, A.M- An evaluation of bone growth into porous high density polyethylene. J Biomed Mater Res 1976; 1 0:33 1-3 1 3.

Klein, C.P.A., Driessen, A.A., and de Groot, K. Biodegradation behaviour of various calcium phophate materials in bone tissues. J Biomed Mater Res 1983; 17:769- 784.

Klein, C.P.A.T., de Blieck-Hogervorst, J.M.A., Wolke, J.G.C., and de Groot, K. Solubility of hydroxylapatite, tricalcium phosphate and tetracalcium phosphate coatings in vitro. Adv. Biomat. 1 990;9:277-282.

Klein, C. A.T., Patka,P., van der Lubbe, H.B.M., Wolke, J.G.C., and de Groot, K. Plasma-sprayed coatings of tetracalcium phophate, hydroxyl-apatite and a-TCP on titanium alloy : An interface study. J Biomed m e r Res 199 1 ;25:53-65.

Lakkakorpi, P.T. and Vaananen, HK. Kinetics of the Osteoclast Cytoskeleton During the Resorption Cycle In Vitro. J Bone Minerd Res 2 991 ;6 (8):8 17-825.

Lane, NE., Kaneps, A.J., Stover, S.M., Modin, G., and -el, D.B. Bone mineral density and turnover following forelimb immobilization and recovery in young adult dogs. Calcif Tissue Inr 1996;59(5):40 1 -406.

Langer, R- and Vacanti, J.P. Tissue Engineering- Science 1993;260:920-926.

Lee, J.and Sukyoung, K. Preparation of macorporous hydroxyapatite Ceramics. Presented at the Fifth World Biomaterials Congress May29-June 2, 1996, Toronto, Canada p.53.

LeGeros, R.Z. Calcium phosphate materials in restoraûve dentistry: a review. A h , Dent Res l988;î: 164- 1 80,

LeGeros, R.Z. Calcium Phosphates in Oral Biology and Medicine. New York, Karger Press 199 1.

Le Geros, R.Z. and Le Geros, J.P (1993). Dense hvdroxva~atite. In An introduction to Bioceramics, edited by L.L. Hench and J.Wilson. World Scientific, Singapore, pp. 139-1 80.

LeGeros, R.Z., and LeGeros, J.P (1 995). Caiciurn Phosphate Biomaterials: f re~aration, Properties. and Biodemdation. In Ency lopedic Handbook of B iomaterials & Bioengineering Part A: Materials Vol. 2 edited by D.L. Wise, D.J. Trantoio, D.E. Altobelli, M.J. Yaszemski, J.D. Gresser and E.R. Schwartz. Marcel Dekker, New York, pp. 1429- 1463.

Lindholm, T. S. and Gao, T.J. Functional carriers of bone morphogenetic protein Ann. Chir. Gynaecol. 1983;82;3-12.

Lyndon, K. L. and William, L.R. (1996) Osteowrosis. In Principles of Bone Biology, edited by J.P. Bilezikian, L.G. Raisz and G.A.Rodan. Academic Press,

pp. 941-950.

Makley, J.T. The use of allografts to reconstnict Uitercalary defects of long bones. CIin Orthop 1985; l97:58-75.

Mainard, D., Galois ,L., Bordji, &, Clement, D., and Delagoutte, I.P. Bone Lngrowth into Porous Ceramics with Different Pore Sizes. Presented at the Fi* World Biomaterials Congress May29-June 2, 1996, Toronto, Canada, p.?

Maniatopoulos, C., Sodek, J., and Melcher, AH. Bone formation in vitro by stroma1 celIs obtained fiom bone marrow of young addt rats. Ce12 Tissue Res. l988;254:3 17-330.

Mankin, H.J., Doppelt, S., and Tomford, W.W. Clinical experience with allograft implantation: the first ten years. Clin Orthop 1 983; 1 74:69-89.

Marks, S.C. and Henney, D.C. (1996) The Structure and Deveio~ment of Bone. In Priaciples of Bone Biology edited by J.P. Bilezikian, L.G. Raisz and G.A.Rodan. Academic Press, pp. 3-1 5.

Martin, I., Quarts, R., Dozin, B., and Cacedda, R. Producing prefabncated tissues and organs via tissue engineering. IEEE Eng Med Biol1997; l6:73-80.

Martin, R.B. Porosity and specific surface of bone- CRC Citical Reviews in Biomedical Engineering 1984: lO(3): 179-222.

Minuth, W.W., Stockl, G., Kloth, S., and Denneitzel, R. Construction of an apparatus for ce11 pemision which enables in vitro experirnents under organotypic conditions. European J Cell Bio l992;57: 132-1 37.

Minuth, W.W., Sittinger, M., and Kloth, S. Tissue engineering: generation of differentiated artificial tissues for biomedical applications. Cell Tissue Res 1998;219:1-11.

Mnaymneh W., Malinin TL, Makley J.T., and Dick H M . Massive osteoarticular allografts in the reconstruction of extremities following resection of tumors not requiring chemotherapy and radiation. Clin Orrhop 1985; 1 97:76-87.

Mors, W. and Kaminski, E. Osteogenic replacment of tricalcium phosphate ceramic implants in the dog palate. Arch Oral Biol 1975;20:365-367.

Muller-Mai, C .M., Voigt, C., and Gross, U.M. Incorporation and degradation of hydroxyapatite implants of different surface roughness and surface structure in bone. Scanning Microscopy 1 990;4(3):6 1 3-624.

Muller-Mai, C.M, Stupp S.I., Voigt, G., and Gross, U.. Nanoapatite and organoapatite implants in bone: Histology an ultrastructure of the interface. J Biomed Mat Res

Murphey, M.D., Sartoris, D.J. and Bramble, J.M- (1992) Radiom~hic assessrnent of bone gi.afts. In Bone Grafts & Bone Substitutes, edidted by M.B. Habal and A.H. Reddi. W.B. Saunders Company, pp. 9-36.

Nade, S. and Burwell, KG. Decalcified bone as a substrate for osteogenesis: an appraisal of the interrelation of bone and marrow in combined gr&. JBone Joint Surg [m 1977;59: 189-196.

Nagase, M., Chen, R-B., Araya, Y., and Nakajima, T. Evaluation of a bone substitute prepared fiom alpha-tricalcium phoshpate and an acid polysaccyaride solution. J Oral Maih$ac Surg 199 1 ;49: 1305- 1309.

Nerry E.B., Lynch, K.L., Huthe, W.M., and Mueller, K.H. Bioceramic implants in surgically produced infiabony defects. J Periodontol 1975;40:668-675.

Nunes, CR-, Simske, S.F., Sachdeva, R., and Wolford, L.M. Long-term ingrowth and apposition of porous hydroxyapatite implants. J Biomed Mater. Res 1997;36:560- 563.

Ogura, M., Sakae, T., and Davies, LE. (1 99 1) Resomtion of calcium hvdroxva~atite substrats bv osteoclast-like cells in vitro. In Biocetamics Vol. 4, edited by W. Bonfield, G.W, Hastings, and K.E. Tanner. Butterworth Heinemann Ltd., London, UK, pp. 121-126.

Ohgushi,H., Goldberg,V.M., and Caplan, A.I. Repari of bone defects with marrow cells and prorous ceramic. Acta Orthop Scand 1989;60(3):334-339.

Ohgushi, H., Okumura, M., and Tamai, S. Marrow ceil induced osteogenesis in porous hydroxyapapatite and tricalcium phosphate: A comparative histomorphometric study of ectopic bone formation. J Biomed Mater Res 1 WOa; 24: 1 563- 1 570.

Ohgushi, H and Okurnura, M. Osteogenic capacity of rat and human marrow cells in porous ceramics. Acta Orthop Scand 1 99Ob;6 1 (5):43 1 -434.

Ohgushi, H., Okumara, M., Yoshikawa, T.,Inouem, K., Senpuku, N., and Tamai, S. Bone formation process in prous calcium carbonate and hydroxyapatite. J Biomed Mater Res, 1992a;26:88S-895.

Ohgushi, H.,Okumum,, M., Masuhara, K., Goldberg, V.M.,Davy, D.T., and Caplan, A.I. (1 992b) Osteogenic Potential of Bone Marrow Sustained bv Porous Calcium Phos~hate C e d c s . In CRC Handbook of Bioactive Ceramics, Vol II, edited by T.Yamamuro, LL Hench and J-Wilson. Boca Raton, Florida: CRC Press, Inc., pp. 229-233.

Ohgushi, H.,Okumura,, M., Masuhara, K., Goldberg, V.M.,Davy, D.T., and Caplan, A.I. (1992~) Calcium Phomhate Block Ceramic with Bone Marrow Cells in a Rat Long Bone Defect- In CRC Handbook of Bioactive Ceramics, Vol II, edited by T.Yamamuro, LL Hench and LWilson. Boca Raton, Florida: CRC Press, Inc., pp. 235-238.

O' Kelly, K.,Tancred, D.,McCormack, B., and Carr, A. A quantitative technique for comparing sythetic porous hydroxyapatite structures and cancellous bone. J Muter Science: Materials in Medicine l996;7:207-2 1 3.

Panda, R., Teung, P., Danforth, S C , and Safari, A. (1 998) Fabrication of calcium phomhate based ceramics with controlled mrositv for bone scaffolds. In Bioceramics Volume 1 1 (Proceedings of the 1 lh International Symposium on Ceramics in Medicine, New York, NY, USA, edited by RZ. LeGeros and J.P. LeGeros. World Scientific Publishg Co. Pte. Ltd, pp..

Petite, H., Kacem, K., and Triffia, J-T. Adhesion, growth and differentiation of hwnan bone marrow stroma1 cells on non-porous calcium carbonate and plastic substrats: effects of dexamethasone and l,25 dihydroxyvitamin D3. J Mater Science: Materials in Medicine 1996:7:665-671.

Pilliar, R.M.., Deporter, D.A-,Watson, P.A., Abdulla, D., Valiquette, N., and Lindsay, D. Quantitative Determination of Relative Movement Effects on Bone hgrowth with Porous-Coated Implants. Fifth World Biomaterials Congress, May 29-June 2, 1 996, Toronto, Canada, p. 26 1.

Pritchard, J.J. (1972) General histolom of bone- In The Biochernistry and Physiology of Bone, 2"d ed. edited by G.H. Boume. Academic Press, New York, pp.1-20.

Puelacher, W.C., Vacanti, Schloo, F.B. and Vacanti, C.A. Femoral shaft reconstruction using tissue-engineered growth of bone. h t J Oral Maxillofac Sug l996;25:223- 228.

Qui, Q., Ducheyne, P., and Ayyaswamy, PS. (1998) Bioceramic hollow micros~heres for 3-D bone tissue formation in rotatinp bioreactors. In Bioceramics Volume 11 (Proceedings of the 1 1' International Symposium on Ceramics in Medicine, New York, NY, USA, November 1998), edited by R.Z. LeGeros and J.P. LeGeros. World Scientific Publishing Co. Pte. Ltd, pp.5 13-5 16.

Rey, C. (1998) Calcium Phos~hates for Medical A~~lication. In Calcium Phosphate in Biolgoical Systems, edited by A. Amjad. Klawer Academic Publishing, pp. 2 17- 251.

Rodan, G.A. Introduction to Bone Biology . Bone 1 992; 1 3 :S3-S6.

Rohanizadeh, R Padrines, M., Bouler, J.M., Couchowel, D., Fortun, Y., and Daculsi, G.

(1998) A~atite vreci~iation after incubation of biphasic calcium ~hosphate ceramic in various solutions: influence of ions and ~roteins. In Bioceramics Volume 1 1 (Proceedings of the 1 lLh International ~ b ~ o s i u m on Ceramics in Medicine, New York, NY), edited by R.Z. LeGeros and J.P. LeGeros. World Scientific Publishing Co Pte- Ltd. pp. 41 1-480.

Rout, P.G.J, Tarrant, S.F., Frame, J. W., and Davies J.E. (1 987). Interactions between primaw bone ce11 cultures and biomaterials. PART 3: A cornparison of dense and macroporous hydroxy apatite. In Biomaterials and Clinical Applications edited by A. Pizzofemto, P.G. Ravaglioli, and A.J.C. Lee. Elsevier, Amsterdam, pp.59 1-596.

Salama, R Xenogeneic bone grafting in humans, Clin Orrhop 1983; 1 74: 1 13-1 2 1 -

Schugens, C., Grandfils, C., Jerome, R., Teyssie, P., Delree, P., Martin, P, Malgrange, B., and Moonen, G. Preparation of a macroporous biodegradable polylactide implant for neuronal transplantation. J Biomed Mater Res l955;29 9 1 1 0: 1349- 1362.

Sittinger, M., Buija, J., Hammer, C., Minuth, W.W., and Burmester, G.R. Tissue engineering and autologous transplant formation: pratical approaches with resorable biomaterials and new ce11 culture techniques. Biomaferiuls 1996; l7:237-242.

Sittinger, M., Schultz, O-, Keyszer, G., Minuth, W. W. and Burmester, G.R. Artificial tissue in perf ' ion culture. Zn J A r t ~ j k a l Organs 1997;20(1):57-62-

Spector, M., Flemming, W.R., and Kreutner, A. Bone growth ïnto porous high-denisty polyethy lene. J Biomed Mater Res Symposium 1 976;7:595-603.

Soueidan, A.,Gan, O.I., Bouler,J.M., .Gouin, F., and Daculsi, G. Biodegradation of Synthetic Biphasic Calcium Phosphate and Biological Calcified Substratum by Cells of Hemopoietic Origin. Cells und Materials 1995;5:(1):3 1-44.

Tampierï, A, Celotti, G., Szontagh, F., and Landi, E. Sintering and characterization of HA and TCP bioceramics with control of their strength and phase purity. J Mater Science: Marerials in Medicine 1997:8:29-37.

Tobin, W.J. The Interna1 Architecture of the Femur and its Clinical Significance. J. Bone Jt Surg 1 955;37-A(l):S7-72.

Toth, J.M. and Lynch, K.L. (1995) Mechanical and biological characterization of calcium phosphates for use as biomaterials. ui Encylopedic Handbook of Biomaterials Br Bioenginee~g Part A: Materials Vol. 2 edited by D.L. Wise, D.J. Trantolo, D.E. Altobelli, M.J. Yaszemski, J.D. Gresser and E.R. Schwartz- Marcel Dekker, New York, pp. 1465- 1499.

Uemura, T., Yoshikawa, T., Ohgushi, H., Tamai, S., and Tateishi, T. (1998) In vivo osteo~enic Dotent of cultured bone/hydrowa~atite constnict: A com~arison with that of marrow celvhvdroxva~atite comwsite. In Bioceramics Volume II (Proceedings of the 11' International Symposium on Ceramics in Medicine, New York, NY), edited by R Z LeGeros and J-P. LeGeros. World Scientfic Pubiishing Co Pte. Ltd. p. 505-508.

Urist, M.R. Bone morphogenic protein: The molecdarization of skeletal system deveiopment. J Biomed Muter Res 1997; 12(3):343-346.

van Blitterswijk, C.A., Grote, J.J., Kuijpers, W., Blok-van Hoek, C.J.G. and Daems, W.T.H. Bioreactions at the tissuehydmxyapaîite interface. Biomaterials l985;6:243-25 1.

van Blittemvijk, C.A- and Grote, JJ. (1989) Biological performance of ceramics during inflammation and infection. In CRC Critical Reviews in Biocompatibility 5: 13- 43.

van Blitterswijk, C.A., Hessehg, S C , Grote , J.J.,Koerten, H.K., and de Groot, K. The biocompatibitity of hydroxyapatite ceramic: A study of retrieved human middle ear implants. J Biomed Mater Res 199 1 ;24:433-543.

Villanueva, A.R., Sypitkowski,, C., and Parfitt, A.M. A new method of identification of cernent lines in undecalcified, plastic embedded sections of bone. Stuin Technology 1 986;6 1 33-88.

Ward, F.O. Outlines of Human osteology. London:Henry Renshaw, 1838.

Whitehouse, W.J and Dyson, E.D. Scanning electron microscope studies of trabecdar bone in the proximal end of the human femur. J Anatomy 1974; 1 18(3):417-444.

Wilkes, R.P and Athanasiou, K.A. The inminc incompressibility of osteoblast-like cells. Tissue Engineering 1 996;2(3): 167- 1 8 1.

Wolford, L.M., Wardrop, R. W ., and Hartog, J.M. Coraliine porous hydroxylapatite bone gral3 substitute in orthognathic surgery . J. Oral MmiIZofac. Surg. 1 987;4S; 1 034- 1042.

Xie, L and Monroe, E-A (1 992). The Hvdrolvsis of Tetracalcium Phosphate and other Calcium Ortho~homhate. In CRC Handbook of Bioactive Ceramics Vol.II, edited by T.Yarnamuro, LL Hench and J.Wilson. Boca Raton, Florida: CRC Press, inc., pp. 3-16.

Yoshikawa, T., Ohgushi, H., Okumura, M., Tamai, S., Dohi, Y., and Moriyana, T. Biochemical and Histological Sequences of Membranous Ossification in Ectopic Site. Calcif Tissue Int 1992; 50: 184-1 88.

Yoshikawa, T., Ohgushi, H., and Tamai, S. intermediate bone forming capability of prefabnacated osteogenic hydroxyapatite. J Biomed Mater Res l996;32:48 1-492.

Yoshikawa, T., Peel., S.A.F., Gladstone, J.R., and Davies, J.E. Biochemical analysis of the response in rat bone marrow ce11 culture to mechanical stimulation Bio- Medical Materials and Engineering 1 997;7:3 69-3 77.

Zhoy H., ChernecS. R., and Davïes, J.E. Deposition of cernent at reversai fines in rat femoral bone. J Bone Miner Res 199499L367-374-

Zygourakis, K. Quantification and regulation of ceil migration. Tissue Engineering l996;î(l): 1 - 16.