Embed Size (px)
Citation preview
Tissue Engineering Bone - Reconstruction of
Critical Sized Segmental Bone Defects in a
Large Animal Model
Johannes Christian Reichert, MD
Faculty of Built Environment and Engineering, School of Engineering
Systems, Queensland University of Technology
Thesis submitted for:
Doctor of Philosophy (PhD)
2010
I
Keywords
Bone, segmental defect, tibia, tissue engineering, scaffold, tricalcium-
phosphate, polycaprolactone, bone morphogenetic protein, osteoblasts,
mesenchymal stem cells
II
III
Abstract
Currently, well established clinical therapeutic approaches for bone
reconstruction are restricted to the transplantation of autografts and
allografts, and the implantation of metal devices or ceramic-based implants
to assist bone regeneration. Bone grafts possess osteoconductive and
osteoinductive properties, their application, however, is associated with
disadvantages. These include limited access and availability, donor site
morbidity and haemorrhage, increased risk of infection, and insufficient
transplant integration. As a result, recent research focuses on the
development of complementary therapeutic concepts. The field of tissue
engineering has emerged as an important alternative approach to bone
regeneration. Tissue engineering unites aspects of cellular biology,
biomechanical engineering, biomaterial sciences and trauma and
orthopaedic surgery. To obtain approval by regulatory bodies for these novel
therapeutic concepts the level of therapeutic benefit must be demonstrated
rigorously in well characterized, clinically relevant animal models.
Therefore, in this PhD project, a reproducible and clinically relevant, ovine,
critically sized, high load bearing, tibial defect model was established and
characterized as a prerequisite to assess the regenerative potential of a
novel treatment concept in vivo involving a medical grade polycaprolactone
and tricalciumphosphate based composite scaffold and recombinant human
bone morphogenetic proteins.
IV
V
Table of contents
Keywords I
Abstract III
Table of contents V
List of illustrations and diagrams XI
List of abbreviations XVII
Statement of original authorship XIX
Acknowledgments XXI
Introduction 1
Chapter I – Preclinical models for segmental bone defect research
5
- Clinical background 7
- Introduction 9
- Definition of critical-sized defect 9
- Large animal models in bone defect research 11
- Tibial fracture models 15
- Tibial segmental defect models 19
- Summary 33
Chapter II – Ovine bone and marrow derived progenitor cells: Isolation, Characterization, and Osteogenic Potential
37
- Introduction 39
- Materials and Methods 41
- Isolation of ovine MPC and OB 41
- Flow cytometric analysis 42
- Cell proliferation assay 43
- CFU-F clonogenic assay 43
VI
- 2D differentiation in vitro 44
- Dynamic cell culture 44
- Alkaline phosphatase activity 45
- Aliyarin red staining 45
- Wako HR II calcium assay 46
- X-ray photoelectron spectroscopy 47
- Immunohistochemistry 47
- Total RNA isolation, primer design and qRT-PCR 48
- 3D cultures 49
- Scanning electron microscopy 50
- Confocal laser microscopy 50
- In vivo transplantation studies 51
- !CT analysis 52
- Histology 52
- Image analysis 53
- Statistical analysis 53
- Results 54
- MPC show a higher proliferation rate than OB in vitro 54
- MPC and OB exhibit a similar phenotype 55
- Clonogenic efficiency of MPC and OB 56
- 2D differentiation potential of ovine MPC and OB in vitro 56
- Static vs. dynamic culture 60
- 3D differentiation potential of MPC and OB in vitro 65
- Differentiation potential of MPC and OB in vivo 67
- Discussion 70
- Conclusion 76
VII
Chapter III – Ovine bone and marrow derived progenitor cells and their
Potential for Scaffold based Tissue Engineering
Applications in vivo
77
- Introduction 79
- Materials and Methods 81
- Isolation of ovine MPC and OB 81
- Scaffold fabrication and cell seeding 81
- Cell sheet fabrication 82
- In vivo transplantation studies 83
- BrdU labelling of cells 84
- Biomechanical testing 85
- !CT analysis 86
- Immunohistochemistry 86
- Histochemistry 87
- Tartrate resistant acid phosphatase (TRAP) staining 88
- Detection of BrdU-labelled cells 88
- SEM and EDX 89
- Image analysis 89
- Statistical analysis 90
- Results 91
- Biomechanical testing 91
- !CT analysis 91
- Histology 93
- SEM and EDX 107
- Discussion 108
- Conclusion 114
VIII
Chapter IV – Establishment of a Preclinical Ovine Model for Tibial
Segmental Bone Defect Repair by Bone Tissue
Engineering Methods
115
- Introduction 117
- Regulatory framework 119
- Intramedullary nail versus plate fixation versus external
fixator
124
- Road map to establish a preclinical model for segmental bone
defect research
128
- Pilot study limited contact locking compression plate
(LC-LCP)
129
- Finite element modelling
131
- Implant testing 132
- Pilot study dynamic compression plate 134
- Summary 137
Chapter V – Reconstructing large segmental bone defects in an ovine
model by tissue engineering methods
139
- Introduction 141
- Materials and Methods 145
- Scaffold fabrication and preparation 145
- Anaesthesia and pre-operative treatment 147
- Defect model 146
- Harvest of autologous cancellous bone graft 150
- Experimental groups 151
- Euthanasia 151
- Radiographic analysis 152
- Computed tomography 152
- Biomechanical testing 154
IX
- !CT analysis 155
- Histology 157
- Statistical analysis 157
- Results 158
- Animal model 158
- X-ray analysis 158
- Computed tomography 160
- Biomechanical testing 165
- !CT analysis 168
- Histology 173
- Discussion 176
- Summary 185
Conclusions and Recommendations 187
Bibliography 191
X
XI
List of illustrations and diagrams
Chapter I
Figure 1: PMMA sections of ovine tibial bone demonstrating secondary
osteon formation Table 1: Factors influencing the size of a critical sized defect Table 2: Bone biomechanical properties of different species Table 3: Overview of segmental bone defect studies Table 4: Comparison of animal models for fracture and segmental bone
defect research Table 5: Summary of human and large animal bone properties
Chapter II
Figure 1: Mesenchymal progenitor cell and osteoblast isolation and
expansion Figure 2: Uncoated and collagen I coated polycaprolactone tricalcium-
phosphate scaffold Figure 3: Ovine mesenchymal progenitor cell and osteoblast proliferation
over seven days in monolayer Figure 4: Surface antigen expression of ovine mesenchymal progenitor
cells and osteoblasts Figure 5: Clonogenic efficiency of ovine mesenchymal progenitor cells
and osteoblasts Figure 6: Alizarin red, osteocalcin, and type I collagen staining of
mesenchymal progenitor cell and osteoblast monolayers after 28 days of osteogenic induction
Figure 7: Quantification of alizarin red incorporated in mesenchymal
progenitor cell and osteoblast monolayer cultures after 28 days of osteogenic culture
Figure 8: Quantification of alkaline phosphatase activity at day 14 and 28
of osteogenic culture in monolayer
XII
Figure 9: Quantitative RT-PCR for osteogenic markers after 28 days of culture under osteogenic conditions in monolayer
Figure 10: Alkaline phosphatase activity under static and dynamic
monolayer culture conditions Figure 11: Alizarin red quantification at day 14 of static and dynamic
osteogenic monolayer cultures Figure 12: Calcium amounts in dynamic cultures after 14 days Figure 13: XPS analysis of ovine bone and Thermanox coverslips Figure 14: XPS analysis of extracellular matrix synthesized by ovine
mesenchymal progenitor cells and osteoblasts Figure 15: Proliferation of mesenchymal progenitor cells and osteoblasts
under static and dynamic conditions Figure 16: Scanning electron microscopy, FDA-PI and Phalloidin-DAPI
staining of mesenchymal progenitor cells and osteoblasts cultured on mPCL-TCP scaffolds for 28 days
Figure 17: MicroCT analysis of mesenchymal progenitor cell and
osteoblast 3D cultures on collagen I coated polycaprolactone tricalcium-phosphate scaffolds
Figure 18: Histology, histomorphometry and microCT analysis of
transplanted in vivo specimens after eight weeks
Chapter III
Figure 1: Schematic illustrating the in vitro generation of a transplantable
tissue engineered construct Figure 2: Immunohistochemical staining for BrdU of an ovine
mesenchymal progenitor cell monolayer culture Figure 3: Compressive stiffness values of subcutaneously transplanted
tissue engineered constructs after eight weeks Figure 4: 3D microCT reconstructions of subcutaneously transplanted
tissue engineered constructs after eight weeks Figure 5: Bone volume fractions and bone mineral density of
transplanted tissue engineered constructs after eight weeks as determined by microCT analysis
XIII
Figure 6: H&E staining on paraffin sections of cell free mPCL-TCP scaffolds that were transplanted alone, in combination with fibrin glue or with fibrin glue containing rhBMP-7
Figure 7: H&E staining on paraffin sections of mesenchymal progenitor
cell or osteoblast seeded mPCL-TCP scaffolds that were transplanted in combination with or without rhBMP-7
Figure 8: Von Kossa/van Gieson staining on PMMA sections of cell free
mPCL-TCP scaffolds that were transplanted alone, in combination with fibrin glue or with fibrin glue containing rhBMP-7
Figure 9: Histomorphometric analysis of mineralized tissue within
transplanted tissue engineered constructs Figure 10: Histomorphometric analysis of neovascularisation within
transplanted tissue engineered constructs Figure 11: Von Kossa/van Gieson staining on PMMA sections of
mesenchymal progenitor cell or osteoblast seeded mPCL-TCP
scaffolds that were transplanted in combination with or without
rhBMP-7
Figure 12: High magnification of PMMA sections stained for von
Kossa/van Gieson showing differences in morphology of bone lining cells
Figure 13: Immunohistochemistry for osteocalcin on paraffin sections Figure 14: Alcian blue staining on paraffin sections demonstrating
glycosaminoglycan deposits Figure 15: Immunohistochemistry for type II collagen on paraffin sections Figure 16: Histochemical staining for tartrate resistant acid phosphatase
on paraffin sections illustrating osteoclast activity Figure 17: Immunohistochemical staining and histomorphometry for BrdU
labelled cells on paraffin sections Figure 18: Backscattered scanning electron microscopy and Energy
dispersive X-ray spectroscopy
XIV
Chapter IV
Figure 1: Flow chart describing a road map to establish a critically sized defect model
Figure 2: Schematic of commonly applied methods for segmental defect
fixation Figure 3: Image showing a segmental tibial defect of 2 cm length
stabilized with a LC-LCP Figure 4: Image of different implants chosen for biomechanical testing
and set up of the four point bending test Figure 5: Equivalent bending stiffnesses of tested implants Figure 6: Image illustrating a segmental tibial defect of 2 cm length
stabilized with a DCP, postoperative radiograph and x-ray after 12 week
Table 1: Advantages and disadvantages of different fixation devices
Chapter V
Figure 1: 3D microCT reconstruction of a cylindrical mPCL-TCP scaffold
of 3 height and 2cm diameter Figure 2: Image series demonstrating the application of and OP-implant
to a cylindrical mPCL-TCP scaffold Figure 3: Intraoperative images of the critical size defect creation in an
ovine tibia Figure 4: Image series illustrating the harvesting procedure of autografts
from the iliac crest Figure 5: 3D CT reconstruction of a 3 cm tibial defect overlayed with a
developed CT scoring system Figure 6: DICOM image of an intact ovine tibia (axial view)
Figure 7: Embedding procedure for torsional testing Figure 8: Schematic illustrating the method of calculation for the polar
moment of inertia Figure 9: Illustration of the cutting planes for histological sectioning
XV
Figure 10: Radiographs demonstrating bone formation within defects of
the different experimental groups after 12 weeks
Figure 11: Bar graphs demonstrating union rates and determined CT
scores for the different experimental groups
Figure 12: 3D CT reconstructions of representative specimens of each
experimental group
Figure 13: Box plots demonstrating the total bone volumes and newly
formed bone in the cortical region determined by quantitative
analysis of CT scans
Figure 14: Bone volumes determined for the marrow and external callus
regions determined by quantitative analysis of CT scans
Figure 15: Box plot illustrating torsional moment values measured for the
different experimental groups
Figure 16: Box plot showing torsional stiffness values calculated for the
different experimental groups
Figure 17: MicroCT sections and 3D microCT reconstructions of
representative samples of the different experimental groups
Figure 18: Box plots demonstrating the volume of newly formed bone
within the defects determined by quantitative microCT analysis
Figure 19: Bone volume distribution along the z axis of the 3 cm defects
Figure 20: Box plots showing the tissue mineral density of the newly
formed bone within the 3 cm defects
Figure 21: Box plots illustrating the trabecular thicknesses for proximal,
medial and distal defect portions and thickness distribution
along the z axis.
Figure 22: Box plots demonstrating the polar moment of inertia calculated
for the 3 cm defect regions of the different experimental groups
Figure 23: Histological sections of the entire 3 cm defects of each group
stained for Safranin O/von Kossa and Movat’s pentachrome
XVI
Figure 24: Higher magnification images of histological defect sections
stained for Movat’s pentachrome showing the different
composition of tissue formed within the 3 cm defects
Figure 25: High magnification images of PMMA sections stained for
Movat’s pentachrome illustrating the different degrees of bone
maturation within the different experimental groups
XVII
List of abbreviations
2D Two dimensional 3D Three dimensional ABG Autologous bone graft ALP Alkaline phosphatase AO Arbeitsgemeinschaft osteosynthese ARC Australian Research Council bFGF Basic fibroblast growth factor BMP Bone morphogenetic protein BrdU 5-bromo-2-deoxyuridine BSA Bovine serum albumin BV Bone volume CAD Computer aided design CD Cluster of differentation CFU Colony forming unit CO2 Carbon dioxide CSD Critical sized defect DAPI 4',6-diamidino-2-phenylindole DCP Dynamic compression plate DMEM Dulbecco’s modified eagle media DNA Deoxyribonucleic acid ECM Extracellular matrix EDTA Ethylene-di-amine-tetra-acetic acid EDX Energy dispersive X-ray
spectroscopy EU European union f frequency FACS Fluorescence activated cell sorting FBS Foetal bovine serum FDA Fluorescein diacetate FDA Food and drug administration FDM Fused deposition modelling FEM Finite element modelling FGF Fibroblast growth factor FITC Fluorescein isothiocyanate GAGs Glycosaminoglycans GDF Growth differentiation factor H2O water HCl Hydrogen chloride HE Haematoxylin-eosin HGF Hepatocyte growth factor IGF-1 Insulin like growth factor 1 ISO International organization for
standardization LC-DCP Limited contact dynamic
compression plate LC-LCP Limited contact locking compression
plate LISS Less invasive stabilization system
XVIII
MPC Mesenchymal progenitor cell mPCL-TCP Medical grade poly-caprolactone
tricalcium-phosphate MSC Mesenchymal stem cell MVF Mineralized volume fraction (!)CT (Micro) computed tomography NaOH Sodium hydroxide NOD/SCID Nonobese diabetes/severe combined
immunodeficiency OB Osteoblast OC Osteocalcin OP-1 Osteogenic protein 1 PBS Phosphate buffered saline PC-Fix Point contact fixator PCL Poly-caprolactone PDGF Platelet derived growth factor PDLA Poly-D-lactic acid PEG Polyethylene glycol PGA Poly-glycolic acid PI Propidium iodide PLA Poly-lactic acid PLLA Poly-L-lactic acid PMMA Poly-methyl-methacrylate pMOI Polar moment of inertia PRP Platelet rich plasma rh Recombinant human RNA Ribonucleic acid RT-PCR Real time polymerase chain reaction SD Standard deviation SEM Scanning electron microscopy TCP Tricalcium-phosphate TEC Tissue engineered construct TGF Transforming growth factor TRAP Tartrate resistant acid phosphatase UHN Universal humeral nail UTN Universal tibial nail VEGF Vascular endothelial growth factor XPS X-ray photoelectron spectroscopy
XIX
Statement of original authorship
“The work contained in this thesis has not been previously submitted to meet
requirements for an award at this or any other higher education institution. To
the best of my knowledge and belief, the thesis contains no material
previously published or written by another person except where due
reference is made.”
Friday, 18 June 2010
XX
XXI
Acknowledgements
I would like to gratefully acknowledge Prof. Dietmar W. Hutmacher for his
friendship and enthusiastic supervision of my work. I thank Prof. Michael
Schütz and Prof. Georg Duda for stimulating, critical discussions, their
support and intellectual input, and Dr. Martin Wullschleger, Dr. Siamak
Saifzadeh and the team of the QUT Medical Engineering Research Facility
for their assistance with the animal surgeries. I’m grateful to Dr. Maria Ann
Woodruff and Dr. Devakara Epari for their friendship, advice and support in
matters of histology, finite element modelling, biomechanics, image analysis,
manuscript writing and proof reading.
I’d further like to thank the members of the IHBI Regenerative Medicine
Group and my fellow research students for companionship, inspiration and
advice.
Finally, I’d like to express gratitude to my wife Verena for her love and
support, sacrifice, patience and proofreading and to my parents and siblings
for support, understanding, patience and encouragement over the years.
XXII
1
Introduction
Bone as such displays a high intrinsic regenerative potential. Consequently,
the majority of fractures and bone defects heal spontaneously. This healing
process recapitulates pathways of embryonic development following complex
yet well orchestrated biological patterns.
Improved surgical techniques and biologically favourable implant designs as
well as novel peri-operative management strategies have significantly
decreased undesirable treatment complication rates. However, various
factors such as a compromised wound environment and biomechanical
instability can result in large defects with impaired healing capacity. In many
aspects, these cases pose a major challenge and dramatically influence
patients’ quality of life. The treatment options are then mainly restricted to
bone graft transplantations (autograft, allograft) or segmental bone transport.
These grafts, however, are limited in availability, face integration related
problems, carry the risk of infection, and are associated with donor site
morbidity. Bone transport on the other hand is a long-lasting, inconvenient
procedure with recurrent pin track infections as a frequent complication.
As a result, recent research efforts have focused on the development and
application of bone graft substitutes and the concept of tissue engineering
has emerged as it unites aspects of cellular biology, biomechanical
engineering, biomaterial sciences, and trauma and orthopaedic surgery.
Tissue engineering generally involves the association of growth factors
and/or cells with a naturally derived or synthetically manufactured,
mechanically supporting scaffold to produce a three-dimensional,
2
implantable construct. Complying with these principles, it was the overall aim
of the present study to develop a bone graft substitute consisting of a
suitable scaffold and a bone growth stimulating agent that results in
comparable or even better bone healing when compared to the standard
cancellous autograft transplantation.
The performance of such bone graft substitutes is usually evaluated in large
animal models as these models simulate human in vivo conditions as closely
as possible. However, most of the preclinical animal models are not well
described, defined or standardized as summarized in Chapter I, which
provides comprehensive information on the advantages and disadvantages
of the different animal models described in the literature. Over the last ten
years, sheep have become increasingly popular as a model species to
investigate bone remodelling and turnover and it was therefore decided to
assess the performance of the developed bone graft substitute in an ovine
critically sized tibial defect. Surprisingly, the molecular and cellular events
surrounding osteoneogenesis in these models have only been scarcely
investigated. Knowledge of cellular behaviours and molecular processes,
and comparison with data available on human conditions, however, provides
essential information pertaining to the validity of an animal model.
It has been demonstrated that - amongst others - mesenchymal progenitor
cells and osteoblasts exhibit distinct intrinsic osteogenic characteristics, are
capable of synthesizing bone-like extracellular matrix, and therefore play a
major role in bone formation in humans.
Hypothesizing that ovine marrow derived cells of seven to eight year old
sheep are equivalent to those previously described for adult humans,
3
populations of marrow cells derived from the iliac crest and cells from
compact bone were isolated, characterized, and compared regarding
phenotype, genotype, and osteogenic potential in vitro using techniques
already established for human mesenchymal progenitor cells as described in
Chapter II of this study.
The processes of de novo bone formation in vivo are partly attributed to
members of the transforming growth factor-" superfamily, specifically the
bone morphogenetic proteins (BMPs). In particular BMP-7 was shown to
have not only osteoinductive but also angiogenic effects. However, the
molecular and cellular events surrounding osteoneogenesis remain poorly
understood with respect to neovascularisation, and the recruitment and
differentiation of osteogenic precursor cell populations. The potential of ex
vivo expanded ovine marrow and bone derived cells to produce tissues with
properties consistent with those of mature bone and the influence of
recombinant human BMP-7 on bone formation was therefore assessed in
vivo in a small animal model of ectopic bone formation. It was hypothesized
that bone cell origin, ossification type, and degree of vascularisation and
bone neoformation is dependent on the nature and commitment of
transplanted cells as well as supplemented growth factors such as BMP-7.
The results obtained from these experiments are summarized in Chapter III
of this study.
Translations from bench to bedside in the field of bone tissue engineering
are still infrequent which might be related to difficulties in integrating
individual technical discoveries in model tissue engineering systems, in
manufacturing scale up, in funding, and in regulatory approval. Translating a
4
tissue engineering concept to a clinical setting requires a rigorous
demonstration of the level of therapeutic benefit in clinically relevant animal
models. As preclinical models published are difficult to reproduce, effort was
subsequently concentrated on the establishment of a well characterized,
standardized preclinical animal model. The approach of how to translate
tissue engineering of bone from bench to bedside is illustrated in Chapter IV.
In this chapter a road map is presented which describes the validation of the
functionality of a highly load-bearing large animal model to study the
regeneration of critically sized segmental bone defects.
The last chapter of the present study finally summarizes the large animal
study designed to investigate the effects of recombinant human BMP-7 in
combination with a composite scaffolds built by computer aided design using
rapid prototyping technologies on bone healing in an ovine, tibial, critically
sized, segmental bone defect.
5
Chapter I
Preclinical animal models for segmental bone defect research
6
7
Clinical Background
The vast majority of fractures and bone defects heal spontaneously. This
process is stimulated by well balanced biological and micro-environmental
conditions. Newly introduced, improved surgical techniques, implant designs
and peri-operative management strategies have procured better treatment
outcomes of complex fractures and other skeletal defects caused by high
energy trauma, disease, and tumours [1-6]. However, a compromised wound
environment, biomechanical instability and other factors can result in large
defects with limited intrinsic regeneration potential [7]. Such defects pose a
major surgical, socio-economical and research challenge, and highly
influence patients’ quality of life [8, 9].
Cancellous bone fractures of the proximal humerus, distal radius, or the tibial
plateau can lead to bone impaction and consequently defect formation after
reduction [4]. The tibial diaphysis, however, represents the most common
anatomic site for segmental bone defects since it is devoid of muscle
coverage on its anteromedial surface [8]. This poor soft tissue coverage both
increases the risk of bone loss and complicates treatment [8].
Over the years, bone autografts have advanced as the “gold standard”
treatment to augment and accelerate bone regeneration [1, 2, 10-16]. The
application of autografts, however, is associated with considerable negative
side effects. Graft harvest leads to prolonged anaesthetic periods and
requires personnel [12, 14, 17]. Often, insufficient amounts of graft can be
obtained while the access to donor sites is limited [12, 13, 18, 19]. Donor site
morbidity (persistent pain and haemorrhage) is common, the risk of infection
is increased, and the transplanted bone is predispositioned to failure [4, 12,
8
13, 20]. Graft failures usually result from incomplete transplant integration,
particularly in large defects [14]. In addition, graft devitalisation due to
insufficient graft vascularisation and subsequent resorption processes can
lead to decreased mechanical stability [21].
The transplantation of vascularised autografts is time consuming and
technically demanding. Allografts and xenografts carry the risk of immune-
mediated rejection, graft sequestration, and transmission of infectious
disease [9, 22-28]. The dense nature of cortical bone allografts impedes
revascularization and cellular invasion from host sites following implantation
[18]. This limited ability to revascularize and remodel is believed to account
for an allograft associated failure rate of 25% and a complication rate of 30-
60% [18, 29]. In addition, the maintenance of bone banks is associated with
considerable operating expenses.
A technique introduced to avoid graft integration-related difficulties known as
the “Ilizarov technique”, involves the osteotomy of bone combined with
distraction to stimulate bone formation. This procedure has been applied
successfully to treat large bone defects, infected non-unions, and limb length
discrepancy [30]. However, the Ilizarov technique is a long-lasting procedure,
inconvenient for the patient [31, 32], and recurrent pin track infections are a
frequent complication [24, 33].
In order to avoid the limitations associated with the current standard
treatment modalities for segmental bone deficiencies, research efforts have
focused on the use of naturally derived and synthetic bone graft substitutes
during the past decades.
9
More recently, the concept of tissue engineering has emerged as an
important approach to bone regeneration related research. Tissue
engineering unites aspects of cellular biology, biomechanical engineering,
biomaterial sciences, and trauma and orthopaedic surgery. Its general
principle involves the association of cells with a natural or synthetic
supporting scaffold to produce a three-dimensional, implantable construct.
Introduction
To biomechanically simulate human in vivo conditions as closely as possible,
and to assess the effects of implanted bone grafts and tissue engineered
constructs on segmental long bone defect regeneration, a number of large
animal models have been developed. However, most of the preclinical
models published are not well described, defined or standardized. In 2008,
the Journal of Bone and Joint Surgery published a number of review papers
on preclinical models in fracture healing and non-unions [34]. However,
these articles provide only rudimentary information on how to establish
relevant segmental bone defects in a preclinical large animal model. Hence,
the following chapter provides detailed, comprehensive information on the
advantages and disadvantages of the different published animal models.
Definition of a Critical-Sized Bone Defect
It has been postulated that an experimental osseous injury inflicted to study
bone repair mechanisms needs to be of dimensions to preclude spontaneous
healing [35]. Therefore, the non-regenerative threshold of bone was
determined in different research animal models inducing so-called “critical
10
sized” defects. These critical sized defects are defined as “the smallest size
intraosseous wound in a particular bone and species of animal that will not
heal spontaneously during the lifetime of the animal” [29, 36, 37] or as a
defect which shows less than ten percent bony regeneration during the
lifetime of an animal [37].
Table 1: Factors influencing the size of a critical defect
The minimum size that renders a defect ‘‘critical’’ is not well understood. For
practical reasons, it has been defined as a segmental bone deficiency of a
length exceeding 2-2.5 times the diameter of the affected bone [24, 33].
Results of various animal studies suggest that critical sized defects in sheep
however, could even be approximately three times the diameter of the
corresponding diaphysis [33]. Nevertheless, a critical defect in long bones
cannot simply be defined by its size, but also depends on the species
phylogenetic scale, anatomic defect location, associated soft tissue, and
biomechanical conditions in the affected limb as well as age, metabolic and
systemic conditions, and related co-morbidities affecting defect healing
(Table 1)[24, 36].
Factors determining a CSD [24] [29, 36]
o Age o Species phylogeny o Defect size o Anatomic location o Bone structure and vascularisation o Presence of periosteum o Adjacent soft tissue o Mechanical loads and stresses on the limb o Metabolic and systemic conditions o Fixation method/stiffness o Nutrition
11
Large animal models in bone defect research
Animal models in bone repair research include representations of normal
fracture-healing, segmental bone defects, and fracture non-unions, in which
regular healing processes are compromised in the absence of a critical-sized
defect site [38]. In critical-sized segmental defect models, bridging of the
respective defect does not occur despite a sufficient biological
microenvironment as critical amounts of bone substance are removed. In
contrast, in true non-unions, deficient signalling mechanisms, biomechanical
stimuli or cellular responses may prevent defect healing rather than the
defect size.
When selecting a specific animal species as a model system, a number of
factors need to be considered. With comparison to humans, the chosen
animal model should clearly demonstrate both significant physiological and
pathophysiological analogies in respect to the scientific question under
investigation. Moreover, it must be manageable to operate and observe a
multiplicity of study objects over a relatively short period of time [39-41].
Further selection criteria include costs for acquisition and care, animal
availability, acceptability to society, tolerance to captivity, and ease of
housing [42].
Over the last decades, several publications have described dogs as a
suitable model for research related to human orthopaedic conditions [43]. It
was found that dogs closely resemble humans with regards to bone weight,
density and bone material constituents such as hydroxyproline, extractable
proteins, IGF-1, organic, inorganic and water fraction although clear
differences in bone microstructure and remodelling have been described [44,
12
45]. While the secondary structure of human bone is predominantly
organized into osteons, osteonal bone structure in dogs is limited to the core
of cortical bone, whereas in areas adjoining the periosteum and endosteum
mainly laminar bone is found as characteristic for large, fast-growing animals
[46]. It has been reported that generally, higher rates of trabecular and
cortical bone turnover can be observed in dogs compared to humans [47]
and differences in loads acting on the bone, as a result of the dog’s
quadrupedal gait must also be taken into consideration as well. Various
biomechanical properties as described in table 2.
Bone biomechanical properties
Cortical bone
Trabecular bone
Dog Humerus (bending) E: 2.66 (GPa) UStress 193.23 (MPa) [48]
Femur E: 209 (MPa) UStress: 7.1 (MPa) Tibia E: 106-426 (MPa) UStress: 2-24 (MPa) [40]
Sheep Femur (compression) E:19.3 (GPa) UStrain: 0.019 [49] [50]
Tibia E: 1192 (MPa) Ustress: 21.4 (MPa) [51]
Goat Tibia (bending)
E: 278.08 (MPa) Bending strength: 46.24 (MPa) [12]
Femur
E: 399-429 (MPa) UStress: 14.1-23.5 (MPa) Tibia E: 532-566 (MPa) UStress: 24.7-26.1 (MPa)
Pig
Femur E: 14.6 (GPa) (plexiform bone) 8.3 (GPa) (Haversian bone) [52]
Femur E: 5900 (MPa) [40]
Human
Femur (compression) E: 14.7-19.7 (GPa) UStrain: 167-215 (MPa) Tibia (compression) E: 24.5-34.3 (GPa) UStrain: 183-213 (MPa) [40]
Femur E: 298 (MPa) UStrain: 5.6 (MPa) Tibia E: 445 (MPa) UStrain: 5.3 (MPa) [40]
Table 2: Bone biomechanical properties of different animal species and humans.
13
A review article by Neyt states that between 1991 and 1995 11% of
musculoskeletal research was undertaken in dogs, results that are confirmed
by Martini et al. who found that, between 1970 and 2001, 9% of orthopaedic
and trauma related research used dogs as animal models for orthopaedic
and trauma related research [43, 53]. Recently, the use of dogs as
experimental models has significantly decreased mainly due to ethical
issues, although between 1998 and 2008, still approximately 9% of articles
published in prominent orthopaedic and musculoskeletal journals described
dogs as animal models for fracture healing research [34].
Mature sheep and goats possess a bodyweight comparable to adult humans
and long bone dimensions enabling the use of human implants [54]. The
mechanical loading environment occurring in sheep is well understood [55,
56]. The loading of the hind limb bones, forces and moments, is roughly half
of that determined for humans during walking. Since no major differences in
mineral composition [57] are evident and both metabolic and bone
remodelling rates are akin to humans [58], sheep are considered a valuable
model for human bone turnover and remodelling activity [59]. Bone histology
however reveals some differences in bone structure between sheep and
humans. In sheep, bone consists principally of primary bone structure [60] in
comparison with the largely secondary, haversian bone composition of
humans [61]. Furthermore, the secondary, osteonal remodelling in sheep
does not take place until an average age of 7-9 years (Fig. 1)[54]. Although a
significantly higher trabecular bone density and greater bone strength are
described for mature sheep when compared to humans, the trabecular bone
14
in immature sheep is weaker, has a lower stiffness and density, a higher
flexibility due to higher collagen content [51], and shows comparable bone
healing potential and tibial blood supply [62].
Fig. 1: Ground and polished bone sections from the tibia of a 7 year old sheep
embedded in poly-methyl-methacrylate (PMMA) and stained with Toluidine blue (A)
and Movat pentachrome (B). The images show secondary bone with clearly
distinguishable osteones (arrows). Secondary osteone formation can only be
observed in sheep older than 7 years and makes the ovine secondary bone
structure comparable to human findings. Bar = 0.05 mm.
In a variety of study designs, pigs are considered the animal of choice and
were - despite their denser trabecular network [63] - described as a highly
representative model of human bone regeneration processes in respect to
anatomical and morphological features, healing capacity and remodelling,
15
bone mineral density and concentration [44, 64]. However, pigs are often
neglected in favour of sheep and goats given that the handling of pigs has
been described as rather intricate [54]. Furthermore, the short length of the
tibiae and femora in the pig might bring about the need for special implants,
as one cannot use implants designed for human use.
Tibial fracture models
Animal fracture models have been widely investigated to identify and further
characterize physiological and pathophysiological processes surrounding
fracture healing of long bones. One of the most important elements in the
study of fracture healing or fixation is the establishment of standardized
methods to create reproducible fractures. Although a substantial number of
articles on fracture models in animals and treatment options have been
published over the last decades, only few publications describe the actual
infliction of a fracture by trauma rather than the creation of a bony defect < 3
mm size by osteotomy, which is generally accepted as an alternative since it
is problematic to standardize. In 1988, Macdonald et al. [65, 66] reported a
device for the reproducible creation of transverse fractures in canine tibiae
utilizing a three-point bending technique.
Similarly, to compare the effects of reamed versus unreamed locked
intramedullary nailing on cortical bone blood flow, Schemitsch et al. created
a standardized spiral fracture by three-point bending with torsion in a
fractured sheep tibia model [67, 68], a method also described by Tepic [69]
to establish a standardized oblique fracture in sheep tibiae in order to
compare healing in fractures stabilized with either a conventional dynamic
16
compression plate (DCP) or an internal point contact fixator (PC-Fix). A
minimally invasive approach to create a multifragmental fracture in the sheep
femur (classification by the Association for the Study of Internal Fixation, AO
type 32-C), in which the bone was weakened by two short, transverse
anterior osteotomies and bi-cortical drill holes created through small
incisions, has recently been described by Wullschleger et al. (unpublished
data). The insertion of two chisels and one blade bar were then used to
initiate cracks connecting both the osteotomies and the drill holes, thereby
creating a standardized multifragmental fracture. This technique could easily
be adopted when establishing standardized tibial fractures as well.
Fracture models of osteotomized long bones have been well characterized
over the years in different large animal species. A number of publications
have described fracture models in dogs since the dog, beside pigs, is
considered the most closely related model for research of human
orthopaedic conditions. The effect of bending stiffness of external fixators on
the early healing of transverse tibial osteotomies was described in a canine
model by Gilbert [70]. Tiedemann et al. assessed densitometric approaches
to measure fracture healing in 6 mm tibial segmental defects and single-cut
osteotomy defects in adult mongrel dogs [71]. Bilateral tibial transverse
osteotomies were performed with a 2 mm gap by Markel et al. to quantify
local material properties of fracture callus during gap healing [72]. To
compare the dosage-dependent efficacy of recombinant human bone
morphogenetic protein-2 (rhBMP-2) on tibial osteotomy healing, adult female
dogs underwent right midshaft tibial osteotomies with a 1 mm gap. The
17
operated bones were stabilized using type I external fixators [73]. In a similar
study by Edwards, bilateral tibial osteotomies were performed to evaluate the
capacity of a single percutaneous injection of rhBMP-2 delivered in a rapidly
resorbable calcium phosphate paste (alpha-BSM) to accelerate bone-healing
[74]. The effect of shock wave therapy on acute tibial fractures was studied
by Wang et al. in adult dogs after creation of bilateral tibial osteotomies with
a 3 mm defined fracture gap [75]. Similar models were also described by
Hupel to compare the effects of unreamed and reamed nail insertion [76].
Jain et al. [77] investigated whether or not the limited contact design of the
low contact dynamic compression plate (LC-DCP) provided advantages over
the dynamic compression plates (DCP) regarding their influence on cortical
bone blood flow, biomechanical properties, and remodelling of bone in
segmental tibial fractures. Nakamura [78] also evaluate the effects of
recombinant human basic fibroblast growth factor (bFGF) on fracture healing
in beagle dogs.
As previously mentioned, mature sheep and goats possess a bodyweight
similar to adult humans, show no major differences in bone mineral
composition with similar metabolic and bone remodelling rates, and therefore
are considered a valuable model for human bone turnover and remodelling
activity often used in fracture research. In the period between 1990 and
2001, sheep as an animal model were used in 9-12% of orthopaedic
research, compared to only 5% between 1980 and 1989 [43]. Over the last
ten years numbers of studies utilizing sheep and goats as animal models
have increased to 11-15% [34].
18
The significance of postoperative mechanical stability for bony repair of a
comminuted fracture was investigated in a sheep study comparing four
commonly applied operative methods of fracture stabilization. In this study, a
triple-wedge osteotomy of the right sheep tibia was performed [79]. Using a
standard osteotomy of the ovine tibia stabilised by an external skeletal
fixator, Goodship et al. elucidated the influence of fixator frame stiffness on
bone healing rates [80]. Wallace et al. [81] used a similar model to
investigate serum angiogenic factor levels after tibial fracture. Likewise,
transverse mid-diaphyseal osteotomies with an interfragmentary gap of 3
mm, as an experimental fracture model in sheep, were used to assess
fracture repair processes [82-85]. To validate the principle of external fixation
dynamization to accelerate mineralized callus formation by in vivo
measurements of callus stiffness, transverse fractures with an
interfragmentary gap of 3 mm width were created in the mid third of the tibial
diaphysis [86]. Hantes et al. investigated the effect of transosseous
application of low-intensity ultrasound on fracture-healing in a midshaft
osteotomy sheep model [87]. Epari et al. were the first authors to report on
the pressure, oxygen tension and temperature in the early phase of callus
tissue formation of six Merino-mix sheep that underwent a tibial osteotomy to
model fracture conditions [88]. In this study, the tibia was stabilized with a
standard monolateral external fixator. It was found that the maximum
pressure during gait increased from three to seven days. During the same
interval, there was no change in the peak ground reaction force or in the
interfragmentary movement. Oxygen tension in the haematoma was initially
high post surgery and decreased steadily over the first five days. The
19
temperature increased over the first four days before reaching a plateau on
day four.
Mechanical strain during callus distraction is known to stimulate
osteogenesis. It is, however, unclear whether this stimulus can enhance the
healing of a fracture without affecting bone length. Just recently, Claes et al.,
reported the acceleration of fracture healing by a slow temporary distraction
and compression of a diaphyseal osteotomy [89] in an ovine, mid-diaphyseal
osteotomy fracture model of the right tibia, stabilized by external fixation.
Tibial segmental defect models
In order to ascertain whether newly developed bone graft substitutes or
tissue engineered constructs (TEC) comply with the requirements of
biocompatibility, mechanical stability and safety, the materials must be
subject to rigorous testing, both in vitro and in vivo. To extrapolate results
from in vitro studies to in vivo patient situations, however, is often difficult.
Therefore, the application and systematic evaluation of new concepts in
animal models is often an essential step in the process of assessing newly
developed bone grafts prior to clinical use. To simulate human in vivo
conditions as closely as possible, a variety of large critical sized tibial defect
models - mainly in sheep - have been developed over the past decade in
order to investigate the influence of different types of bone grafts on bone
repair and regeneration. Critical sized segmental defects in long bones are
usually defined by multiplying the diaphyseal diameter by 2.0-2.5 [24, 33].
Interestingly, the method of ostectomy may influence the study outcome.
20
Kuttenberger et al. could show that CO2-laser osteotomy impaired the
adjacent bone less than oscillating saw osteotomy [90].
To evaluate the effects of different bioceramics on bone regeneration during
repair of segmental bone defects Gao et al. [91] implanted biocoral and
tricalcium phosphate cylinders (TCP) in sheep tibial defects of 16 mm length.
The defects were stabilized medially using two overlapping contoured auto-
compression plates of 4 mm thickness (8 and 6 holes) and cortical screws.
When compared to TCP, a significant increase in external callus formation
and callus density was seen with the biocoral implants after three weeks and
an increase of torque capacity, maximal angle of deformation, and energy
absorption could be measured after 12 weeks while microscopically
osseointegration appeared better. However, in his study, Gao used both
male and female animals with a relatively large variation in body weight. Both
factors, gender and body weight are known to have an influence on bone
regeneration due to effects on both the biomechanical environment and
hormonal feed-back control mechanisms. Hence, variations in sex and body
weight should be avoided. The defect fixation method used in this study can
most likely be interpreted as a means to countervail bending forces on the
implant after earlier failures. However, defect fixation by overlapping plates is
not necessarily lege artis and has never been introduced and applied
clinically. Therefore, a thicker and hence stiffer plate should have been
chosen instead.
Den Boer et al. reported a new segmental bone defect model where a 30
mm segmental defect was inflicted on sheep tibiae and stabilized by an
interlocking intramedullary nail (custom made AO unreamed humeral nail).
21
X-ray absorptiometry was applied to quantify healing [59]. Groups of this pilot
study included untreated controls and autografts. After 12 weeks, despite
higher bone mineral density in the autograft group, no significant difference
in torsional strength and stiffness could be revealed. Since 33% of the
control animals showed sufficient bridging of the defect, it needs to be
questioned if the authors succeeded in establishing a reliable non-union
model. Removal of the periosteum or a larger defect site might have been
beneficial. In a subsequent study, the authors described the fabrication of
biosynthetic bone grafts and their application in the very same animal model
[4]. The five treatment groups included empty controls, autografts,
hydroxyapatite alone, hydroxyapatite combined with recombinant human
osteogenic protein I (rhOP-1), and hydroxyapatite with autologous bone
marrow. At 12 weeks, healing of the defect was evaluated radiographically,
biomechanically and histologically and revealed a two-fold higher torsional
strength and stiffness for animals treated with autograft and hydroxyapatite
plus rhOP-1 or bone marrow. Since healing was only evaluated after 12
weeks, no conclusions could be drawn regarding the process of healing or
bone remodelling. The mean values of both combination groups were
comparable to those of autografts. A higher number of defect unions was
described when hydroxyapatite plus rhOP-1 was applied rather than
hydroxyapatite alone. Analysing this study, it has to be taken into account
that animals treated with hydroxyapatite and bone marrow were of a different
breed with a higher average body weight. Animals were held at a different
holding facility and accustomed to unequal forage all of which possibly could
have influenced study outcomes.
22
Bone healing in critical sized segmental diaphyseal defects in sheep tibiae
was also investigated by Gugala et al. [3, 37]. Defects were bridged with a
single porous tubular membrane or with anatomically shaped porous double
tube-in-tube membranes. Membranes with different pore structures were
applied alone and/or in combination with autologous bone graft. The
diaphyseal defects were 40 mm in length and stabilized with a bilateral AO
external fixator. Operated animals were six to seven years of age. Of the six
treatment groups defect healing could only be observed in groups where the
defect was filled with autologous cancellous bone graft and covered with a
single perforated membrane or where the bone graft was administered in a
space between a perforated internal and external membrane. The authors
partly contributed the healing effect to their membrane system; however a
control group, where autologous bone graft is administered without any
membrane was not described. It could also be criticized that post surgery
animals were suspended in slings over the entire experimental period
preventing the animals from sitting and therefore getting up, thus not
reflecting normal, physiological, load bearing conditions.
Wefer et al. [92] conducted a study to develop and test a scoring system
based on real-time ultrasonography to predict the healing of a bone defect.
Defects were filled with a porous hydroxyapatite bone graft substitute or
cancellous bone graft from the iliac crest and stabilized by anterolateral plate
osteosynthesis. After sacrifice, tibiae were tested for torsion to failure. The
results were then correlated with radiographic and ultrasound scores. Sheep
with ceramic implants that developed non-unions showed a significantly
lower score than sheep with sufficient implant integration. A significant
23
correlation between these scores and the biomechanical results was found.
However, although the authors describe their 20 mm defect as a critical sized
model, no control group was included for proof of principle. Hence, the
critical nature of the defect in this study can be questioned.
The effects of new resorbable calcium phosphate particles and paste forms,
which harden in situ after application, on bone healing were investigated by
Bloemers et al. [93]. A 30 mm segmental tibial defect was established and
fixed by a custom made AO unreamed interlocking titanium tibial nail. Twelve
weeks after defect reconstruction, radiological, biomechanical, and
histological examinations were performed. Radiographically, the resorbable
paste group performed better than all other groups. Biomechanical tests
revealed a significantly higher torsional stiffness for the resorbable calcium-
phosphate paste group in comparison with autologous bone. The study
indicated that new calcium phosphate based materials might be a potential
alternative for autologous bone grafts in humans. As with several other
studies, animals of a minimum age of two years with a significant variation in
body weight were used in this study. As mentioned before, it must be
considered that secondary osteonal bone remodelling in sheep does not
occur until an age of seven to nine years. Therefore, it might be difficult to
extrapolate results from this study for applications in adult human patients as
human bone primarily undergoes secondary osteonal bone remodelling.
Insulin-like growth factor I (IGF-1) exerts an important role during skeletal
growth and bone formation. Therefore, its localized delivery appears
attractive for the treatment of bone defects. To prolong IGF -1 delivery,
Meinel et al. entrapped the protein into biodegradable poly(lactide-co-
24
glycolide) microspheres and evaluated the potential of this delivery system
for new bone formation in a non-critical 10 mm segmental tibia defect [94].
The defect was stabilized using a 3.5 mm 11 hole DCP. Administration of
100 !g of IGF-1 in the microspheres resulted in bridging of the segmental
defect within 8 weeks. To avoid excessive load on the operated limbs and
fracturing of the freshly operated tibial defects, the animals were
accommodated in a suspension system for a period of 4 weeks
postoperatively thus preventing physiologic-like biomechanical conditions.
When interpreting data published in this study, it must be taken into account
that the close position of the screws to the defect proximally and distally, and
the obvious fact that the screws at the defect site had not been inserted at a
defined angle might have influenced and biased the outcomes.
In a 48 mm tibial defect model in sheep, ceramic implants of 100% synthetic
calcium phosphate multiphase biomaterial were evaluated [95]. The defect
was stabilized with a 4.5 mm neutralizing plate. Although not reported by the
authors, one can observe bent plates and axial deviations in presented x-ray
and CT images, hence, from a clinical point of view, it must be concluded
that the chosen fixation in that model was insufficient. The presented x-ray
series of the two year animal suggests that the internal fixation device had
been explanted 12-14 weeks post surgery, a fact not described and
explained by the authors. Assuming recovery and bone regeneration without
any complications, in human patients, internal fixation devices would usually
not be removed until 12-18 months post implantation. Good integration
between the ceramic implants and the adjoining proximal and distal bone
ends was observed. A progressive increase in new bone formation was seen
25
over time, along with progressive resorption of the ceramic scaffold. Based
on x-ray analysis, at the one-year time-point, approximately 10% to 20% of
the initial scaffold substance was still present, and after two years it was
almost completely resorbed. The authors state that approximately 10-20% of
the periosteum were deliberately left in situ as a source of osteogenic cells.
However, one might conclude that this procedure appears to be rather
difficult to standardize in order to develop a reproducible model.
Another study using an ovine segmental defect model investigated the
influence of recombinant human transforming growth factor 3 (rhTGF"-3) on
mechanical and radiological parameters of a healing bone defect [96]. In four
to five year old sheep, an 18 mm long osteoperiosteal defect in the tibia fixed
with a unilateral external fixator was treated by rhTGF"-3 delivered by a
poly(L/DL-lactide) carrier, with the carrier only, with autologous cancellous
bone graft, or remained untreated. Weekly in vivo stiffness measurements
and radiological assessments were undertaken as well as quantitative
computed tomographic assessments of bone mineral density in four week
intervals. The follow up of the experiment was twelve weeks under partial
weight bearing since animals were kept in a support system to prevent
critical loads on the fixator and its interface to bone thus not reflecting
physiological loading conditions. The 18 mm defect size described as
spontaneously non-healing, might not have been sufficient to establish a
non-union model in a fully weight bearing biomechanical environment. In the
bone graft group, a significantly higher increase in stiffness was observed
than in the PLA/rhTGF"-3 group and a significantly higher increase than in
the PLA-only group. The radiographic as well as the computer tomographic
26
evaluation yielded significant differences between the groups, indicating the
bone graft treatment performed better than the PLA/rhTGF"-3 and the PLA-
only treatment.
Sarkar et al. assessed the effect of platelet rich plasma (PRP) on new bone
formation in a 25 mm diaphyseal tibial defect in sheep [97]. The defect was
stabilized with a custom-made intramedullary nail (stainless steel, diameter
proximal 12 mm, distal 10 mm) with two locking screws each proximal and
distal. To reduce stress at the screw/bone interface, a custom made
stainless steel plate was additionally applied medially therefore choosing an
unconventional fixation method not applied clinically. However, no reasoning
for the additional medial plating was provided by the authors. Defects were
treated with autologous PRP in a collagen carrier or with collagen alone. A
control group to demonstrate the critical nature of the defect was also not
included. After 12 weeks, the explanted bone specimens were quantitatively
assessed by X-ray, computed tomography (CT), biomechanical testing and
histological evaluation. Bone volume, mineral density, mechanical rigidity
and histology of the newly formed bone in the defect did not differ
significantly between the PRP treated and the control group, and no effect of
PRP upon bone formation was observed. The aforementioned studies are
summarized in table 3.
27
Author Animal age (years)
Defect size (mm)
Follow-up (months)
Fixation Animal housing
Support
Gao et al., 1997 a 16 4
Overlapping autocompression plates, 8 and 6 holes, 4 mm thickness
a a
DenBoer et al.,
1999/03 a 30 3
Custom-made AO unreamed nail (Synthes)
a a
Gugala et al.
1999/02 6-7 40 4
AO bilateral external fixator
Single boxes
Suspension slings
Wefer et al., 2000
# 2 20 12 Anterolat. plate osteosynthesis (not specified)
a a
Bloemers et al., 2003
# 2 30 3
AO unreamed tibial nail (Synthes)
6-8 animals in a 60 m
2 cage
a
Meinel et al.,
2003 a 10 5
3.5 mmm DCP, 11 holes a a
Mastrogiacomo
et al, 2006 2 48 12
4.5 mm plate (not specified), 10-12 holes
Single boxes
Fibre glass cast
Maissen et al., 2006
4-5 18 3 Unilateral external fixator Single
boxes Custom-made support system
Sarkar et al., 2006
5.5-7 25 3
Custom-made intramedullary nail plus medial stainless steel plate
Single boxes
a
Tyllianakis et
al., 2007 1-2 10,20,30 4
Universal Humeral Nail (UHN, Synthes)
Single boxes for 3 days post surgery
a
Liu et al., 2007 a 26 8 Circular external fixator Single
boxes a
ainformation not provided by the authors
Table 3: The table lists a selection of publications on segmental bone defect studies
in sheep tibiae and summarizes animal age, selected defect size, defect fixation,
animal housing as well as supportive devices. The majority must be considered
short term studies where no complete bone remodelling can be expected during the
experimental period. In many cases, authors fail to report important information
concerning animal age, housing and supportive devices.
In 2007 Tyllianakis [98] determined the size of a bone defect that can be
restored with one-stage lengthening over a reamed intramedullary nail in
sheep tibiae. Sixteen adult female sheep were divided into four main groups:
a simple osteotomy group (group I) and three segmental defect groups (10,
20, and 30 mm gaps, groups II-IV). One intact left tibia from each group was
also used as the non-osteotomized intact control group (group V). In all
cases, the osteotomy was fixed with an interlocked Universal Humeral Nail
28
(UHN-Protek-Synthes). Healing of the osteotomies was evaluated after 16
weeks by biomechanical testing. The examined parameters included
torsional stiffness, shear stress, and angle of torsion at the time of fracture.
The regenerate bone in the groups with 10 and 20 mm gaps was of
considerable mechanical properties. Torsional stiffness in these two groups
was nearly equal and values represented about 60% of the stiffness
observed in the simple osteotomy group. Gradually decreasing stiffness was
observed as the osteotomy gap increased. No significant differences were
found among the angles of torsion at fracture for the various osteotomies or
the intact bone.
Teixeira et al. treated tibial segmental defects of 35 mm size in both male
and female sheep aged four to five months. Considering the age of the
animals and the preservation of the periosteum, the critical size of this defect
can be questioned and results cannot necessarily be extrapolated to adult
humans, as described correctly by the authors. An empty control group was
not included in the experiment. The bone defects in the diaphysis of the right
hind limb were stabilized with a titanium bone plate (103 mm in length, 2 mm
thickness, and 10 mm width) combined with a titanium cage. As reported by
the authors, plate bending occurred in 42% of the animals and was partly
attributed to the connection of the titanium cage to the plate. However, it
appears that the bending of the plate was rather a result of insufficient
thickness of the fixation device. The titanium cages were either filled with
autologous cortical bone graft or with a composite biomaterial consistent of
inorganic bovine bone, demineralised bovine bone, a pool of bovine bone
morphogenetic proteins bound to absorbable ultra-thin powdered
29
hydroxyapatite and bone-derived denaturized collagen. Bone defect healing
was assessed clinically, radiographically and histologically. Titanium cages
might keep implanted scaffolds and biomaterials in place initially and
biomechanically support defect fixation, however, it must be taken into
consideration that – since titanium is not resorbable – the cages might hinder
complete bone remodelling in the long run.
Radiographic examination showed initial formation of periosteal callus in both
groups at osteotomy sites, over the plate or cage 15 days postoperatively. At
60 and 90 days callus remodelling occurred. Histological and morphometric
analysis 90 days post surgery showed that the quantity of implanted
materials still present were similar for both groups while the quantity of newly
formed bone was less (p=0.0048) in the cortical bone graft group occupying
51 +/- 3.46% and 62 +/- 6.26% of the cage space, respectively [99].
Recently, Liu et al. reported on the use of highly porous beta-TCP scaffolds
to repair goat tibial defects [12]. In this study, fifteen goats were randomly
assigned to one of three groups, and a 26 mm-long defect at the middle part
of the right tibia in each goat was created and stabilized using a circular
external fixator. In Group A, a porous beta-TCP ceramic cylinder seeded with
osteogenically induced autologous bone marrow stromal cells was implanted
in the defect of each animal. In Group B, the same beta-TCP ceramic
cylinder without any cells was placed in the defect. In Group C, the defect
was left untreated. In Group A, bony union could be observed by gross view,
X-ray and micro-computed tomography (!CT) detection, and histological
observation at 32 weeks post-implantation. The implanted beta-TCP
scaffolds were almost completely replaced by host bone. Bone mineral
30
density in the repaired area of Group A was significantly higher than in Group
B, in which scant new bone was formed without complete resorption of the
beta-TCP after 32 weeks. Moreover, the tissue-engineered bone of Group A
had similar biomechanical properties as the contralateral tibia in terms of
bending strength and Young's modulus. In Group C, little or no new bone
was formed and non-union occurred, demonstrating the critical nature of the
defect.
To investigate the effect of chondroitin sulphate on bone remodelling and
regeneration, Schneiders et al. [100] created a 30 mm tibial mid-diaphyseal
defect site and reconstructed it using hydroxyapatite/collagen cement
cylinders. Defect stabilization was achieved by insertion of a universal tibial
nail (UTN, Synthes, Bochum). To insert the scaffold to the defect, the authors
had to use a second operative aditus mid-diaphyseally. The published data
suggest problems with defect fixation not only due to reported implant
failures but also to clearly evident signs of locking bolt loosening, poor
contact between bone and nail, and the proximal nail end extending into the
articular space. Moreover, it can be supposed that either the insertion of the
nail or undesired movement of the loosened nail has caused damage to the
implants. When interpreting the presented data, it also has to be taken into
account that obviously no fabrication method has been described to reliably
reproduce implants of corresponding geometrical shape.
Rozen et al. investigated whether blood-derived endothelial progenitor cells
promote bone regeneration once transplanted into an ovine, critical sized,
tibial defect [101]. Cells were isolated and expanded in vitro. 2 x 107 cells in
0.2 ml saline were transplanted two weeks after a 32 mm defect had been
31
created (n=7). Defect fixation was achieved by a 4.5 mm stainless steel plate
with four screws each proximally and distally. In the control group (n=8) 0.2
ml saline were injected. Defect bridging was observed in six out of seven
animals in the experimental group. In the control group, five out of six defects
analysed via !CT showed discontinuous (two animals) or minute bridging
(three animals) as stated by the authors. No reference to the remaining three
animals of the control group was found throughout the manuscript.
Therefore, the critical nature of the defect has to be questioned. Not
resecting the periosteum and screw loosening as clearly evident in the
published x-ray images might have contributed to defect bridging in the
control group.
The regenerative capacity of xenogenic human and autologous ovine
mesenchymal progenitor cells was assessed by Niemeyer et al. in an ovine
critical-size defect model [102]. Human and ovine MSC from bone marrow,
were cultured on mineralized collagen and implanted into a 30 mm-long
sheep tibia bone defect (n=7). Unloaded mineralized collagen served as
control. The 30 mm mid-diaphyseal defects were fixed with a seven hole LC-
LCP (Synthes) and a carbon fibre reinforced poly-ether-ketone plate
(snakeplate, Isotec AG, Altstätten, Switzerland). Animals were kept in
suspending slings for eight weeks post surgery. Nevertheless, implant failure
occurred in one animal requiring immediate euthanasia. Wound healing
related problems were reported for another animal.
In the same study, bone healing was assessed up to 26 weeks. Presence of
human cells after xenogenic transplantation was analysed using human-
specific in situ hybridization. Radiology and histology demonstrated
32
significantly better bone formation after transplantation of autologous ovine
MSC on mineralized collagen compared to unloaded matrices and to the
xenogenic treatment group. No local or systemic rejection reactions could be
observed after transplantation of human MSC although the presence of
human MSC could be demonstrated.
The rapid progression of bone graft research and the great number of novel
developments must be supported by systematic assessment based on
clinical practicability and experience, the knowledge of basic biological
principles, medical necessity, and commercial practicality. From the current
literature review, it can be concluded, that in the majority of the mentioned
studies, follow up periods, which in most cases don not exceed six months,
are not suitable to evaluate long-term effects of bone substitutes and
scaffolds on bone regeneration and remodelling, and to determine in vivo
resorption kinetics of the respective biomaterial. Variations in defect sizes
and methods of defect fixation as well as postoperative treatment and
management concepts make it difficult to compare studies and draw reliable
conclusions. The modifications of commercially available fixation devices and
supporting systems to prevent peak loads from acting on implants suggest
the occurrence of implant failures usually expected early after surgery. As a
result, most experimental settings do not reflect the actual clinical conditions
faced and impede the extrapolation of results.
33
Summary
The reconstruction of large bone segments remains a significant clinical
problem. Large bone defects occur mainly as a result of extensive bone loss
due to pathological events such as trauma, inflammation, and surgical
treatment of tumours. Present therapeutic approaches include the application
of bone graft transplants (autografts, allografts, xenografts), as well as
implants made of different synthetic and natural biomaterials or segmental
bone transport. However, no existing therapy has been proven to be fully
satisfactory. As a result, a large number of research groups, that work on the
development of new bone grafting materials, carriers, growth factors, and
tissue engineered constructs for bone regeneration, are interested in
evaluating their concepts in reproducible large segmental defect models. The
optimization of cell-scaffold combinations and locally or systemically active
stimuli will remain a complex process characterized by a highly
interdependent set of variables with a large range of possible variations.
Consequently, these developments must be nurtured and evaluated by
clinical experience, knowledge of basic biological principles, medical
necessity, and commercial practicality. The area of bone tissue engineering,
which has its main focus on the development of bioactive materials, depends
on the use of animal models to evaluate both experimental and clinical
hypotheses. To tackle major bone tissue engineering problems, researchers
must rely on the functional assessment of biological and biomechanical
parameters of generated constructs. However, to allow comparison between
different studies and their outcomes, it is essential that animal models,
fixation devices, surgical procedures and methods of taking measurements
34
are standardized to achieve the accumulation of a reliable data pool as a
base for further directions to orthopaedic and tissue engineering
developments.
35
Mo
de
l a
pp
lica
tio
n
o
Stu
dy o
f n
orm
al fr
actu
re h
ea
ling
o
D
rug
de
live
ry t
o f
ractu
re s
ite
s
o
Eff
ects
of
-
Dru
gs
-
Gro
wth
ho
rmo
ne
s
-
An
gio
ge
nic
fa
cto
rs
-
LA
SE
R
on
fra
ctu
re h
ea
ling
o
E
ffe
ct
of
fixa
tio
n m
eth
od
s o
n p
eri
oste
al, c
ort
ica
l a
nd
so
ft t
issu
e b
loo
d f
low
[1
03
] o
E
ffe
ct
of
typ
e a
nd
rig
idity o
f fixa
tio
n o
n f
ractu
re
he
alin
g a
nd
ra
te o
f re
mo
de
llin
g
[10
4]
o
Cre
atio
n o
f n
on
-un
ion
s
[10
5]
o
Cre
atio
n o
f a
n in
fecte
d b
alli
stic w
ou
nd
mo
de
l o
E
va
lua
tio
n o
f va
rio
us a
sse
ssm
en
ts (
e.g
. x-r
ay)
of
fr
actu
re h
ea
ling
o
In
vitro
te
stin
g o
f sp
ina
l fr
actu
re f
ixa
tio
n s
yste
ms
o
Stu
dy o
f in
teg
ratio
n,
de
gra
da
tio
n a
nd
re
mo
de
llin
g
of
bo
ne
su
bstitu
tes
!
Bo
ne
gra
ftin
g
!
Au
tog
raft
!
Allo
gra
ft
!
Xe
no
gra
ft
!
Bio
ma
teri
als
of
na
tura
l a
nd
syn
the
tic o
rig
in
!
De
min
era
lise
d b
on
e m
atr
ix
!
Bio
ma
teri
als
[1
06
] 1
. H
yd
roxya
pa
tite
/tri
ca
lciu
m-p
ho
sp
ha
te
ce
ram
ics
2.
Po
lym
ers
3
. M
eta
ls
4.
Co
mp
osite
s
o
Bo
ne
su
bstitu
tes p
lus a
uto
ge
no
us b
on
e m
arr
ow
o
E
va
lua
tio
n o
f o
ste
og
en
ic p
ote
ntia
l o
f ce
ll-
se
ed
ed
co
mp
osite
im
pla
nts
an
d a
sse
ssm
en
t
of
oste
oin
du
ctive
pro
pe
rtie
s o
f g
row
th f
acto
rs
De
fect
fixa
tio
n
Inte
rna
l o
In
tra
me
du
llary
ro
d/p
in
o
Pla
te a
nd
scre
ws
o
Scre
ws
o
Ce
rcla
ge
wir
es
Ex
tern
al
o
Exte
rna
l fixa
tors
o
C
asts
o
S
plin
ts
No
fix
ati
on
M
ice
an
d r
ats
c
Inte
rna
l o
In
tra
me
du
llary
na
il/p
in
o
Pla
te a
nd
scre
ws
Ex
tern
al
o
Exte
rna
l fixa
tors
Me
tho
ds o
f d
efe
ct
form
atio
n
o
Ma
nu
ally
cre
ate
d f
ractu
res
b
o
Th
ree
-po
int
be
nd
ing
b
o
Gu
illo
tin
e-l
ike
ap
pa
ratu
sb
o
Oscill
atin
g s
aw
, h
igh
sp
ee
d
de
nta
l b
urr
, o
r scis
so
rsb
o
Ba
llistic in
jury
o
Oscill
atin
g s
aw
o
Gig
li’s w
ire
An
ima
l sp
ecie
s a
nd
de
fect
loca
tio
n
Do
ga
o
Fe
mu
r [1
07
] [2
4]
o
Tib
ia [7
2]
[70
] [7
1]
[73
] [7
4]
[76
] [7
7]
[78
] o
R
ad
ius [
10
8]
[10
9]
[11
0]
[11
1]
S
he
ep
a
o
Tib
ia
[89
] [6
8]
[85
] [
79
, 1
12
] [8
0]
[81
] [8
8]
[69
] [1
13
] G
oa
t o
T
ibia
[1
14
] [9
8]
[11
5]
Pig
o
P
elv
is [
11
6]
o
Fe
mu
r [1
17
] o
T
ibia
[1
18
] [1
19
] o
S
pin
e [
12
0]
[12
1]
Do
ga
o
Fe
mu
r [1
07
] [2
4]
o
Tib
ia [7
2]
[70
] [7
1]
[73
] [7
4]
[76
] [7
7]
[78
] o
R
ad
ius [
10
8]
[10
9]
[11
0]
[11
1]
S
he
ep
a
o
Tib
ia
[89
] [6
8]
[85
] [
79
, 1
12
] [8
0]
[81
] [8
8]
[69
] [1
13
] G
oa
t o
T
ibia
[1
14
] [9
8]
[11
5]
Pig
o
P
elv
is [
11
6]
o
Fe
mu
r [1
17
] o
T
ibia
[1
18
] [1
19
] o
S
pin
e [
12
0]
[12
1]
a M
ost com
monly
used
b M
eth
ods to m
imic
accid
enta
l fr
actu
res m
ore
clo
sely
c In
radia
l, u
lnar
or
fibula
r fr
actu
res w
here
additio
nal bony s
upport
is p
resen
t
Ta
ble
4
An
ima
l M
od
els
of
Fra
ctu
res
(Oste
oto
my)
An
ima
l M
od
els
of
Se
gm
en
tal B
on
e
De
fects
(O
ste
cto
my)
Table 4: Comparison of animal models for fracture and segmental bone defect
research
36
Ha
ve
rsia
n
ca
na
l
15
-50
18
-12
0
18
-70
27
-17
0
Dia
me
ter
(µm
) o
f [1
22
]
Ha
ve
rsia
n
sy
ste
m
12
5-1
75
75
-36
0
50
-32
5
18
0-3
25
Bo
ne
c
om
po
sit
ion
[4
2]
Bo
ne
m
ine
ral
de
nsity s
imila
r to
hu
ma
ns
Hig
he
r tr
ab
ecu
lar
bo
ne
de
nsity t
ha
n
hu
ma
ns
(0.6
1g
/cm
3)
Ve
ry s
imila
r to
h
um
an
Ve
ry s
imila
r to
h
um
an
Tra
be
cu
lar
bo
ne
d
en
sity:
0.4
3
g/c
m3
Bo
ne
re
mo
de
llin
gb
[42
]
10
0%
La
rge
r a
mo
un
t o
f b
on
e in
-g
row
th t
ha
n
hu
ma
ns
Sim
ilar
to
hu
ma
ns
Sim
ilar
to
hu
ma
ns
10
-15
%
to 4
0-5
5%
Ad
va
nta
ge
s
[42
] [4
4]
[54
]
o
Tra
ct
ab
le n
atu
re
o
Sim
ilar
bo
ne
min
era
l d
en
sity t
o
hu
ma
ns
o
Do
cile
a
nim
als
o
S
imila
r b
od
y w
eig
ht
to
hu
ma
ns
o
Dim
en
sio
n
of
lon
g b
on
es
su
ita
ble
fo
r h
um
an
im
pla
nts
Mo
re t
ole
ran
t to
am
bie
nt
co
nd
itio
ns
Bo
ne
min
era
l d
en
sity,
an
ato
my,
mo
rph
olo
gy,
an
d
he
alin
g s
imila
r to
h
um
an
s
Dis
ad
va
nta
ge
s
[40
] [5
4]
[12
3]
[42
]
o
Hig
he
r ra
te o
f so
lid
bo
ny f
usio
n w
he
n
co
mp
are
d t
o h
um
an
s
o
Lo
w n
on
-un
ion
ra
tes
o
Eth
ica
l is
su
es a
nd
n
eg
ative
pu
blic
p
erc
ep
tio
n
Sig
nific
an
t in
ter-
an
ima
l va
ria
tio
ns
du
e t
o b
ree
d
div
ers
ity
o
Ag
e-d
ep
en
da
nt
bo
ne
re
mo
de
llin
g
o
Ha
ve
rsia
n
rem
od
elli
ng
at
7-9
ye
ars
of
ag
e (
with
m
ed
ium
-siz
ed
, ir
reg
ula
r ca
na
ls)
Inq
uis
itiv
e a
nd
in
tera
ctive
na
ture
o
Hig
h g
row
th r
ate
s
an
d e
xce
ssiv
e b
od
y
we
igh
t o
D
ifficu
lt h
an
dlin
g
Ap
pli
ca
tio
n
[42
]
Mu
scu
loske
leta
l a
nd
de
nta
l re
se
arc
h
Ort
ho
pa
ed
ic
rese
arc
h
Re
se
arc
h o
n
ca
rtila
ge
, m
en
isci
an
d lig
am
en
tou
s
rep
air
Ort
ho
pa
ed
ic a
nd
d
en
tal stu
die
s
(fe
mo
ral h
ea
d
oste
on
ecro
sis
, fr
actu
res,
bo
ne
in
-g
row
th,
de
nta
l im
pla
nts
)
a N
um
be
r o
f H
ave
rsia
n s
yste
ms p
er
sq
ua
re m
illim
ete
r bA
ve
rag
e w
ho
le b
od
y t
rab
ecu
lar
bo
ne
tu
rno
ve
r p
er
ye
ar
Ma
cro
-str
uc
ture
[1
1]
[12
4]
[12
5]
Fe
mu
r:
Pro
no
un
ce
d c
urv
atu
re a
t d
ista
l th
ird
of
sh
aft
; n
arr
ow
in
th
e m
idd
le
Tib
ia:
Le
ng
th s
imila
r to
sh
ee
p;
pro
xim
ally
co
nve
x
me
dia
lly,
dis
tally
co
nve
x la
tera
lly;
pro
xim
al
$ p
rism
atic,
rem
ain
de
r cylin
dri
ca
l
Fe
mu
r:
Ro
un
de
d (
cylin
dri
ca
l) s
ha
ft;
co
nve
x
do
rsa
lly;
cu
rve
d in
dis
tal %
; re
gu
lar
in
dia
me
ter
Tib
ia:
Ma
jor
we
igh
t-b
ea
rin
g b
on
e o
f cru
s;
lon
g
an
d s
len
de
r; s
ha
ft c
urv
ed
me
dia
lly a
nd
ca
ud
ally
at
ce
ntr
e;
rou
nd
in
mid
dle
, tr
ian
gu
lar
pro
xim
ally
, fla
tte
ne
d c
ran
io-
ca
ud
ally
in
dis
tal th
ird
; m
ed
ial su
rfa
ce
is
su
bcu
tan
eo
us
Alm
ost
ide
ntica
l to
sh
ee
p
Fe
mu
r:
Re
lative
ly w
ide
& m
assiv
e d
iap
hysis
with
4
su
rfa
ce
s
Tib
ia:
Slig
htly c
urv
ed
dia
ph
ysis
, co
nve
x m
ed
ially
o
Ep
iph
ysis
- p
roxim
al a
nd
dis
tal-
sp
on
gy b
on
e
o
Me
tap
hysis
- tr
an
sitio
n z
on
e-s
po
ng
y
bo
ne
dis
tal to
ep
iph
yse
al lin
e
o
Dia
ph
ysis
- s
ha
ft o
f th
e b
on
e –
co
mp
act
bo
ne
Wo
ve
n-f
ibe
red
bo
ne
(p
lexifo
rm b
on
e)
Se
co
nd
ary
oste
on
s (
with
sm
all
ca
na
ls)
incre
ase
in
n
um
be
r w
ith
ag
e
Pri
ma
ry b
on
e s
tru
ctu
re
(ple
xifo
rm c
ort
ica
l b
on
e)
in
yo
un
g s
he
ep
(3
-4 y
ea
rs o
f a
ge
)
No
n-h
om
og
en
eo
usly
d
istr
ibu
ted
Ha
ve
rsia
n
syste
ms
Ple
xifo
rm b
on
e (
oste
on
al
ba
nd
ing
) D
en
se
Ha
ve
rsia
n b
on
e
(with
me
diu
m c
an
als
),
incre
ase
s w
ith
ag
e
Ma
inly
cir
cu
mfe
ren
tia
l la
me
llar
bo
ne
Wo
ve
n-f
ibe
red
bo
ne
(ple
xifo
rm)
form
ed
on
ly in
ra
pid
bo
ne
re
pa
ir a
nd
rem
od
elli
ng
Mo
re H
ave
rsia
n b
on
e t
ha
n
q
ua
dru
pe
ds (
12
.87
vs.
5.5
a)
Ta
ble
5
M
icro
-str
uc
ture
[4
0]
[61
] [1
22
, 1
23
]
Do
g
Sh
ee
p
Go
at
Pig
Hu
ma
n
Table 5: Summary of human and large animal bone properties
37
Chapter II
Ovine bone and marrow derived progenitor cells: Isolation,
Characterization, and Osteogenic Potential
38
39
Introduction
In general, bone displays a high intrinsic regenerative capacity following
insult or disease. Therefore, the majority of bone defects and fractures heal
spontaneously, stimulated by well orchestrated endogenous cell populations
and micro-environmental cues. Improvements in surgical techniques, implant
design and peri-operative management have significantly improved
treatment outcomes of complex fractures and other skeletal defects resulting
from high energy trauma, disease, developmental deformity, revision
surgery, and tumour resection [1-6]. However, a compromised wound
environment, insufficient surgical technique or biomechanical instability can
lead to formation of large defects with limited regeneration potential [7]. Such
defects pose a major surgical, socio-economical and research challenge and
can significantly influence patients’ quality of life [8, 9]. Over the years, bone
grafts have advanced as the “gold standard” treatment for bone
augmentation [1, 2, 10-16]. However, the use of autologous bone is
associated with additional anaesthetic time and personnel required for graft
harvesting [12, 14, 17]. Often, insufficient amounts of graft can be obtained
while access to donor sites is limited [12, 13, 18, 19]. Donor site pain, nerve
damage or haemorrhage can occur while donor bone is predispositioned to
failure [4, 12, 13, 20]. Synthetic and naturally derived bone graft substitutes
have been investigated thoroughly during the past decades to address these
limitations. This research has brought about the rise of tissue engineering as
an important alternative approach to bone related orthopaedic and trauma
research. Orthopaedic research often necessitates the use of animal models.
Therefore, a number of large animal models has been developed to
40
biomechanically simulate human in vivo conditions as closely as possible,
and to assess the effects of implanted bone grafts and tissue engineered
constructs on segmental long bone defect regeneration. In particular, mature
sheep are considered a valuable model for human bone turnover and
remodelling activity since animals of seven to nine years of age show similar
bone structure and composition, possess a bodyweight comparable to adult
humans, and long bone dimensions enabling the use of human implants [34,
40, 42, 54]. However, published fracture and segmental defect models have
often used considerably younger animals [126]. Whilst, sheep are a well
recognized animal model for bone related research, the molecular and
cellular events surrounding fracture and bone defect healing remain poorly
understood with respect to the recruitment and differentiation of osteogenic
precursor cell populations.
In the present study it was hypothesized that ovine marrow derived cells of
seven to eight year old sheep are equivalent to those previously described
for adult humans. The aims were to isolate, characterize and compare
populations of marrow cells derived from the iliac crest with cells from
compact bone. Initial characterization of cell properties, phenotype and
genotype was performed using techniques already established for human
mesenchymal progenitor cells (MPC). The potential of ex vivo expanded
ovine marrow and bone derived cells to produce tissues with properties
consistent with those of mature bone was further assessed in vitro and in
vivo. The purpose of this study was to provide a detailed characterisation of
the ovine segmental bone defect model on a cellular level.
41
Materials and Methods
Isolation of ovine MPC and OB
Ovine osteoblast (OB) explants were obtained from 6-7 year old Merino
sheep undergoing experimental surgery as approved by the animal ethics
committee of the Queensland University of Technology, Brisbane, Australia
(ethics number 0700000915). Compact tibial bone samples were collected
under sterile conditions, minced, washed with PBS (Invitrogen) and vortexed
prior to incubation with 10 ml 0.25% trypsin/EDTA (Invitrogen) for 3 min at
37°C, 5% CO2. After trypsin inactivation with 10 ml low glucose Dulbecco's
Modified Eagle Media (DMEM) containing 10% foetal bovine serum (FBS)
(Invitrogen), samples were washed once with PBS and transferred to 175
cm2 tissue culture flasks (Nunc). Samples were topped-up with 12 ml of
DMEM containing 10% FBS and 1% penicillin/streptomycin. Osteoblast
outgrowth could be observed after 5-7 days (Fig. 1). Cells were expanded to
the second or third passage for subsequent experiments.
Bone marrow aspirates were obtained from the iliac crest under general
anaesthesia. Total bone marrow cells (0.5-1.5 x 107 cells/ml) were plated at
a density of 1-2 x 107 cells/cm2 in complete medium comprising low glucose
DMEM supplemented with 10% FBS, 100 U/ml penicillin and 100 !g/ml
streptomycin. Cells were subsequently plated at a density of 103 cells/cm2.
42
Fig. 1: Cultures of MPC isolated from ovine bone marrow and OB derived from tibial
compact bone (&) after 3, 7 and 14 days in culture. Magnification 10x.
Flow cytometric analysis
Ex vivo expanded populations were used at passage three of culture for
immunophenotypic analysis. Ex vivo expanded MPC or OB were treated with
trypsin/EDTA and resuspended in blocking buffer for 30 min. Individual tubes
containing 1 x 105 cells were incubated with murine monoclonal IgG
antibodies reactive to either ovine CD14, CD31, and CD45 (Serotec), ovine
and human CD29, CD44 (Clone H9H11; Division of Haematology, IMVS,
Adelaide, SA, Australia), and CD166 (BD Biosciences), or isotype matched
controls, 1B5 (IgG1), and 1A6.11 (IgG2b) at a concentration of 10 !g/ml for 1
h on ice. After washing, cells were incubated with secondary detection
reagents, goat anti-mouse IgG-FITC or IgM-FITC conjugated antibodies
(1:50; Southern Biotechnology Associates, Inc., Birmingham, AL) for 45 min
on ice. Following washing, samples were analysed using a Cytomics FC 500
flow cytometry system (Beckman Coulter).
43
Cell proliferation assay
Adherent passage three MPC and OB were seeded in triplicates at 3000/cm2
in flat bottomed 24-well plates (Nunc) and maintained in 1 ml standard
culture medium consisting of low glucose DMEM supplemented with 10%
FBS for 1, 3, 5 or 7 days in a humidified atmosphere (37°C, 5% CO2). At
each time point, cells were washed twice with PBS and stored at -80°C until
analysis. For analysis, samples were digested overnight with 0.5 mg/ml
proteinase K in 1 x Tris-EDTA buffer at 55°C. DNA content for 100 µl of each
sample in triplicate was measured and quantified using a Quant-iT
PicoGreen dsDNA assay kit according to the protocol supplied by the
manufacturer (Invitrogen). An equal volume of the Quant-iT PicoGreen
aqueous working solution was added to each triplicate and incubated for 3
min on a rocking plate. Fluorescence was measured with a Polar Star
Optima plate reader (BMG Labtech, Offenburg, Germany) at an excitation
wavelength of 485 nm and an emission wavelength of 520 nm.
CFU-F clonogenic assay
CFU-F assays were performed as described previously [127] [128]. Briefly,
passage two sheep osteoblasts and bone marrow derived, mesenchymal
progenitor cells were plated at a density of 0.25 x 104 in six-well plates. Cells
were maintained in low glucose DMEM supplemented with 100 U/ml
penicillin G, 100 µg/ml streptomycin, and 10% or 20% FBS respectively, at
37°C with 5% CO2 for 6 days. To enumerate colonies, cultures were washed
with PBS, fixed in ice cold methanol for 15 min, and stained with 0.05% w/v
crystal violet in dH2O for 15 min. Stained aggregates of greater than 50 cells
44
were scored as CFU-colonies under a light microscope (TS100-U, Nikon,
Melville, NY).
2D differentiation in vitro
Passage three MPC and OB were seeded in triplicate into 6-well plates
(Nunc) at a density of 3000 cells/cm2 and expanded in low glucose
DMEM/10% FBS (Invitrogen) until confluent. Osteogenic induction was then
performed over the following 28 days using DMEM/10% FBS supplemented
with 50 µg/ml ascorbate-2-phosphate, 10 mM "-glycerophosphate, 0.1 µM
dexamethasone (Sigma-Aldrich). Controls were cultured in standard
expansion medium (DMEM/10% FBS).
Dynamic cell culture
For dynamic cell culture, MPC and OB were seeded into 6-well plates (Nunc,
Rochester, NY) at a density of 3000 cells/cm2 in triplicates and expanded in
low glucose DMEM/10% FBS (Invitrogen, Carlsbad, CA) until confluent.
MPCs were then differentiated osteogenically with low glucose DMEM/10%
FBS or DMEM/20% FBS each supplemented with 50 µg/ml ascorbate-2-
phosphate, 10 mM "-glycerophosphate, 0.1 µM dexamethasone both under
standard static and dynamic culture conditions on a rocking plate (Bioline
platform rocker, Edwards Instruments, Narellan, Australia; f=0.125 Hz).
45
Alkaline phosphatase activity
At day 14 and 28, ALP enzyme activity was quantified using a colorimetric
assay. Triplicates were washed with PBS, incubated with 0.1% Triton X in
0.2 M Tris buffer at -20°C for 10 min. Cells were harvested and centrifuged at
10000 rpm for 10 min at 4°C and 100 µl of the cell extraction supernatant
was incubated with 125 µl p-Nitrophenylphosphate (1mg/ml) in 0.2 M Tris
buffer (Sigma-Aldrich) in 96 well plates (Nunc) and OD was measured after
30 min at 405 nm in a Polar Star Optima plate reader. ALP activity was
normalized against the sample DNA content determined using a Quant-iT
PicoGreen dsDNA assay kit (Invitrogen).
Alizarin red staining
To determine matrix mineralization, at day 14 and 28, triplicate samples were
washed twice with PBS and fixed with ice cold methanol for 10 min at room
temperature. Samples were then washed twice with ddH2O and incubated
with 1% alizarin red s (Sigma-Aldrich) in ddH2O, pH 4.1 for 10 min with
gentle shaking. After aspiration of the unincorporated dye, samples were
washed three times with ddH2O and air dried. Stained monolayers were
documented using inverted phase microscopy (TS100-U, Nikon, Melville,
NY). For quantification of staining, 800 µl 10% (v/v) acetic acid was added to
each well, and the plate incubated at room temperature for 30 min with
shaking. The monolayer was then scraped from the plate with a cell scraper
and transferred with 10% (v/v) acetic acid to a 1.5 ml microcentrifuge tube.
After vortexing for 30 s, the slurry was overlaid with 500 µl mineral oil
(Sigma-Aldrich), heated to 85°C for 10 min, and transferred to ice for 5 min.
46
The slurry was then centrifuged at 20,000 x g for 15 min and 500 µl of the
supernatant was removed to a new 1.5 ml microcentrifuge tube. Then 200 µl
of 10% (v/v) ammonium hydroxide were added to neutralize the acid.
Aliquots (150 µl) of the supernatant were read in triplicate at 405 nm in 96-
well format using opaque-walled, transparent-bottomed plates (Nunc).
Obtained values were normalized against the DNA content of separate
samples since treatment with acetic acid was expected to cause
denaturation of DNA.
Wako HRII calcium assay
To further analyse the calcium contents in extracellular matrices produced by
ovine MPC or OB, a calcium assay was performed as per the manufacturer’s
instructions (Wako). Briefly, samples were washed with ddH2O and
incubated with 800 !l of 10% acetic acid at room temperature for 30 min.
Monolayers were scraped off, heated to 85˚C for 10 min, transferred to ice
for 10 min, then 200 !l of 10% ammonium hydroxide was added to 500 !l of
each sample. Triplicate 10 !l aliquots were transferred into a 96 well plate, to
which 100 !l of a monoethanolamine buffer, pH 11.0, was added for a 3 min
incubation at 37˚C, followed by 100 !l of o-cresolphthalein complexion and
incubation for 5 min at 37˚C. To generate a standard curve, a dilution series
of Multichem calibrator A (Wako) in 10% acetic acid was used.
Measurements were taken at '=570 nm in a Polar Star Optima plate reader
(BMG Labtech).
47
X-ray photoelectron spectroscopy
For X-ray photoelectron spectroscopy (XPS) cells were seeded on
thermanox coverslips and cultured under osteogenic conditions as described
above. Samples were fixed with glutaraldehyde and dehydrated. XPS
spectra were acquired on a Kratos AXIS Ultra spectrophotometer operating
at a base pressure of 10-9mbar and equipped with a monochromatized Al K(
source. Acquisition was done with an analyser pass energy of 160eV on all
cell culture samples, with an energy of 20eV on native ovine bone and
thermanox coverslips. Samples were investigated with a charge
compensation gun (emission current, 0.15!A).
Immunohistochemistry
For immunohistochemistry, OB and MPC were cultured on Thermanox
coverslips to fit 24-well plates (Nunc). Media was removed; samples were
washed twice with PBS and fixed in 4% paraformaldehyde for 1 h on ice.
Cells were then permeabilized with 0.1% Triton X in PBS for 5 min and
quenched with 0.15 M glycine in PBS for 15 min (Sigma-Aldrich). Samples
were blocked with 1% BSA (Sigma-Aldrich) in PBS for 60 min and incubated
with primary mouse anti-human type I collagen (1:100) (MP Biomedicals,
Irvine, CA) and mouse anti-bovine osteocalcin (1:500) (Takara Bio Inc.,
Japan) antibodies in 1% BSA in PBS for 1 h at room temperature. Samples
were then washed three times with 0.1% BSA in PBS for 5 min each wash
and incubated with a FITC-conjugated goat anti-mouse secondary antibody
(Invitrogen) at a concentration of 1:200 for 30 min. Cover slips were then
48
mounted on glass microscope slides and visualized using a fluorescent
microscope (TE2000-U, Nikon, Melville, NY).
Total RNA isolation, primer design and qRT-PCR
Total RNA was harvested from triplicate wells, from both differentiated and
control cells, on days 7, 14, 21 and 28. Cells were washed twice with PBS
and lysed in 1 ml Trizol reagent (Invitrogen) and RNA isolated following the
manufacturer’s instructions. cDNA was synthesized from 1!g of total RNA
using SuperScript III (Invitrogen) according to the manufacturer’s
instructions. Sheep specific oligonucleotides (Geneworks, TheBarton, SA,
Australia) were designed according to these parameters: 20-30 nt in length;
melting temperature 60ºC, +/- 2ºC; at least one primer spanning an exon
boundary; amplicon length 150 nt, +/- 50 nt; GC content between 40 and
60%; and at the 3’ end a C, G, CG or GC. In those cases where an Ovis
aries mRNA transcripts was not available, a blast search was perform on the
International Sheep Genome Consortium database
(https://isgcdata.agresearch.co.nz/) using the equivalent human mRNA
transcript. To date, the exon boundaries of sheep transcripts have not been
annotated; these boundaries were therefore baseded exon boundary
information for human and mouse mRNA transcripts. Quantitative RT-PCR
was performed on an Applied Biosystems 7900HT FAST Real Time PCR
system (Applied Biosystems, Scoresby, VIC, Australia) using a 384-well
plate layout; templates and reagents were aliquoted using an Eppendorf
5075 epMotion pippeting robot (Quantum Scientific, Murarrie, QLD,
Australia). The reaction volumes per well were as follows: 5 !l 2X SYBR
49
Green (Roche, Castle Hill, NSW, Australia), 1 !l forward and reverse primers
at 1 !M final concentration, 1 !l water, 2 !l cDNA template diluted 1:10 from
stock. The thermo cycling conditions were as follows: 1 cycle of 10 min at
95°C for activation of the polymerase, 40 cycles of 10 sec at 95°C and 1 min
at 60°C for amplification. Dissociation curve analysis was carried out to verify
the absence of primer dimers and/or non-specific PCR products. The
expression of the genes of interest was normalized against the GAPDH
housekeeping gene.
3D cultures
Fused deposition modelling was used to fabricate circular mPCL-TCP
scaffolds of 5 mm diameter and 3 mm thickness. Type I rat tail collagen
(Vitrogen 100, Cohesion, Palo Alto, CA) was lyophilized into the pore space
forming a microporous mesh throughout the polymer (Fig. 2). For 3D
cultures, 120.000 ovine MPC or OB suspended in 60 !l of basal medium
were seeded onto each type I collagen coated mPCL-TCP scaffold and
placed in an incubator. After 1 h, 1 ml of medium was added to each 24-well.
Cell scaffold constructs were cultured in DMEM/20% FBS supplemented with
50 µg/ml ascorbate-2-phosphate, 10 mM beta-glycerophosphate, and 0.1 µM
dexamethasone on a rocking plate (f=0.125 Hz) for up to 4 weeks.
50
Fig. 2: Medical grade polycaprolactone-tricalciumphosphate scaffold (left) coated
with rat tail collagen type I (right) of 5 mm in diameter and 3 mm thickness.
SEM
Cell scaffold constructs were fixed with 3% (v/v) glutaraldehyde in 0.1 M
sodium cacodylate buffer solution (pH 7.3) for 1 h at 4°C. Fixed specimens
were then dehydrated through a series of alcohols; two changes each of
50%, 70%, 90%, and 100% ethanol and were incubated for 10 min between
each change. Specimens were then critical point dried (Denton Vacuum,
Moorestown, NJ) and gold coated in a SC500, Bio-Rad sputter coater (Bio-
Rad) before examination using a FEI Quanta 200 scanning electron
microscope (FEI, Hillsboro, OR).
Confocal laser microscopy
To assess cell viability and morphology of MPC and OB seeded onto type I
collagen coated mPCL-TCP scaffolds, samples were stained with fluorescein
diacetate (FDA) and propidium iodide (PI)(Invitrogen) or rhodamine
conjugated phalloidin and 4',6-diamidino-2-phenylindole (DAPI)(Invitrogen).
For FDA-PI staining, samples were rinsed 3 times with PBS and incubated
with FDA staining solution (2 !g/ml) at 37°C for 15 min in the dark. FDA is a
cell-permeant esterase substrate which is hydrolysed by living/viable cells to
51
give green fluorescence. Samples were then rinsed 3 times with PBS and
incubated with PI staining solution (20 !g/ml) at room temperature for 2 min
in the dark. PI is actively excluded by live cells thus dead cells in a
population are stained red. Samples were again rinsed 3 times with PBS and
visualised with a Leica SP5 confocal microscope (Leica Microsystems
GmbH, Wetzlar, Germany).
For phalloidin-DAPI staining, samples were fixed with 4% paraformaldehyde
for 20 min and permeabilized with 0.2% Triton X-100 in PBS for 20 min at
room temperature with gentle rocking. Samples were then washed twice for
5 min with 1 ml PBS at room temperature, and 700 !l rhodamine-conjugated
phalloidin (0.8 U/ml in 1% BSA in PBS) was added to each sample and
incubated for 1 h at room temperature with gentle rocking. Phalloidin binds to
F-actin of the cytoskeleton. Samples were then washed twice for 5 min with 1
ml PBS at room temperature and nuclei were stained with DAPI staining
solution (1.0 !g/ml in PBS) for 40-50 min at room temperature. Samples
were washed twice for 5 min with 1 ml PBS at room temperature and
visualised with a Leica SP5 confocal microscope.
In vivo transplantation studies
Ovine MPC and OB were seeded onto type I collagen coated mPCL-TCP
scaffolds at a density of 120.000 cells/scaffold and cultured for 4 weeks in
DMEM/20% FBS supplemented with 50 µg/ml ascorbate-2-phosphate, 10
mM "-glycerophosphate, and 0.1 µM dexamethasone on a rocking plate
(f=0.125 Hz). The cell scaffold constructs were then transplanted
subcutaneously into both left and right side pockets formed in the dorsal
52
surface of 10-week-old immunocompromised, male NOD/SCID mice (ARC,
Perth, WA, Australia). Implants were recovered after 8 weeks and fixed in
4% paraformaldehyde.
!CT analysis
Micro CT analysis was performed on both the in vitro constructs and the in
vivo constructs. After 4 weeks of in vitro culture, constructs were carefully
removed from each well and inserted into polycarbonate sleeves for micro-
CT analysis. In vitro mineralization within the constructs was quantified using
a Micro-CT 40 scanner (Scanco Medical, Brüttisellen, Switzerland) at a voxel
size of 6 µm. Samples were evaluated at a threshold of 72, a filter width of
3.0 and filter support of 5.0. In vivo transplanted constructs were scanned at
a voxel size of 16 !m and were evaluated at a threshold of 140, a filter width
of 1.0 and filter support of 2. X-ray attenuation was correlated to sample
density using a standard curve generated by scanning hydroxyapatite
phantoms with known mineral density. Mineralized matrix volume or bone
volume fraction, and mineral density were quantified throughout the entire
construct.
Histology
For histological examination, specimens were fixed in 4% paraformaldehyde,
and dehydrated using an ethanol gradient (30 min in 70%, 1 h in 90%, 95%
and 100% ethanol). The samples were then processed through xylenes for
40 min three times, infiltrated with MMA for 3 h and embedded in MMA
containing 3% PEG. Seven micrometre sections were cut with an
53
osteomicrotome (SM2500; Leica Microsystems, Wetzlar, Germany),
stretched with 70% ethanol onto a polylysine coated microscope slide (Lomb
Scientific), overlayed with a plastic film and slides were clamped together
before being dried for 12 h at 60°C. Sections were then stained using
combined von Kossa and van Gieson [129] stains to visualise the
mineralised bone and connective tissue respectively.
Image analysis
Histology sections were quantified using Image J software to quantify the
amount of mineralisation in a given area of section. Briefly, a JPEG image of
the entire tissue section was selected, converted to grayscale and a scale
bar was calibrated onto the image. The entire tissue section area was then
calculated by segmenting the entire tissue region from the background, and
then measuring the area. Next, only the mineralised (black) area was
segmented from the entire tissue area, and measured. The total mineralised
area was then calculated as a percentage of the total section area. Six
sections were analysed per sample group.
Statistical analysis
Statistical analysis was carried out using the student’s t test and p values <
0.05 were considered significant (SPSS, SPSS Inc., Chicago, Illinois, USA).
54
Results
MPC show a higher proliferation rate than OB in vitro
Cells isolated from ovine bone marrow showed a significantly higher
proliferative potential after 1, 3 and 5 days in culture (p<0.05) when
compared to ovine tibial osteoblasts as represented by the higher DNA
content per 24-well for MPC (Fig. 3). After 6-7 days both MPC and OB
entered a plateau phase indicative for contact inhibition of cells reaching
confluency (p=0.0502).
Fig.3: Proliferative potential of MPC and OB determined by a PicoGreen assay.
When compared to OB, MPC displayed a significantly higher proliferative rate
before reaching confluency between day 5 and 7 (n=3). Values represent the mean
+/- standard deviation.
*
*
*
55
MPC and OB exhibit a similar immunophenotype
Fluorescence-activated cell sorting (FACS) analysis was performed to
characterise the phenotype of ex vivo expanded ovine bone marrow derived
MPC and OB. The two cell populations exhibited similar expression patterns
for CD29 ("1-Integrin), CD44, CD166 (ALCAM) and CD14 (LPS-R). CD44
and CD166 have previously been identified as markers associated with
human bone marrow stromal, adipose, and dental pulp stem cells [127, 130-
132]. Importantly, both populations did not react with hematopoietic markers
CD45 (common leukocyte antigen) and CD31 (PECAM-1, endothelial) (Fig.
4).
Fig. 4: Surface antigen expression for MPC and OB as determined by FACS
analysis. The two cell populations exhibited similar expression patterns for CD29
and CD44. CD166 and CD14 was >50% MPC while only low levels were detected
on OB. Both populations did not react with the hematopoietic markers CD45 and
CD31.
56
Clonogenic efficiency of MPC and OB
The mean frequency of CFU-F derived from marrow aspirates was 2.5 ± 0.5
per 0.25 x 104 mononuclear cells (MNC) in both DMEM/10% FBS and 20%
FBS, with the incidence of CFU-F forming MSC within the marrow MNC
population being approximately 0.08-0.12%. The mean frequency of CFU-F
derived from tibial bone explants was 2 ± 0.6 in DMEM/10% FBS and 1.8 ±
0.7 in DMEM/20% FCS respectively (incidence of colony forming cells 0.04-
0.12%)(Fig. 5).
Fig. 5: Clonogenic efficiency of passage two MPC and OB cultured under defined
media conditions.
2D differentiation potential of ovine MPC and OB in vitro
The potential of bone marrow derived MPC to differentiate into osteoblasts
and of bone derived osteoblasts to secrete a mineralised extracellular matrix
was investigated by culturing cells in the presence of L-ascorbic-2-
phosphate, dexamethasone, and "-glycerophosphate [132]. After 4 weeks of
induction, cultured MPC and OB had formed extensive amounts of alizarin
red-positive mineral deposits throughout the adherent layers. However, OB
consistently formed significantly fewer mineralized nodules (p<0.05)
57
compared to bone marrow derived cells (Fig. 7, Fig. 6 B and H). Extracellular
matrix produced by both MPC and OB stained positive for type I collagen
and osteocalcin (Fig. 6 C-F and I-L). ALP activity measured at day 14 and
day 28 displayed a typical rise-fall pattern [133, 134] and was significantly
increased in osteogenically induced OB and MPC (p<0.05) when compared
to their respective controls (Fig. 8). Under osteogenic conditions a significant
increase in type I collagen expression could be observed over the course of
4 weeks for MPCs whilst no significant increase was found in the control
culture without osteogenic supplements (Fig. 9 A). Osteocalcin expression
was significantly up-regulated around day 14 and further increased towards
the end of week 4 (Fig. 9 C). A significant increase in osteopontin expression
could be detected at day 7 and 14. Osteopontin expression then decreased
towards day 21 and 28 (Fig. 9 E). For OB, no significant changes in type I
collagen expression were found (Fig. 9 B), only a small increase in
osteocalcin expression (Fig. 9 D). Osteopontin levels slightly increased
between day 0 and 7 to further stay on that level (Fig. 9 F).
58
Fig. 6: Alizarin red staining and immunohistochemistry for osteocalcin (OC) and type
I collagen for MPC and OB cultures after 28 days on tissue culture polystyrene.
Under osteogenic conditions (ost), both MPC and OB were found to secrete a
mineralized matrix (B, H) that stained positive for the osteogenic marker protein OC
(D, J) and the extracellular matrix protein type I collagen (F, L). The control cultures
(ctrl) stained negative for both proteins. Magnification 10x.
Fig. 7: Quantification of the incorporated alizarin red s dye. After 28 days of culture
in osteogenic media, MPC cultures showed a significantly higher degree of
mineralization when compared to OB cultures or non-induced controls (ctrl)(n=3).
Error bars represent standard deviations.
59
Fig. 8: Quantification of ALP activity within osteogenically induced (ost) and non-
induced (ctrl) OB and MPC cultures after 14 and 28 days. In osteogenically induced
MPC and OB cultures, ALP activity showed a typical rise and fall pattern and was
significantly higher than for the respective controls (n=3). Error bars represent
standard deviations.
Fig. 9: Quantitative RT-PCR for osteogenic markers. RT-PCR revealed significant
increases in type I collagen and osteocalcin expression over 4 weeks for MPC
under osteogenic conditions (MPC+) and an increase in osteopontin expression
between at day 7 and 14. For OB, no significant changes in type I collagen,
osteopontin and osteocalcin expression were found (n=3).
60
Static vs. dynamic cultures
Culture systems that enhance mass transport and deliver controlled
mechanical stimuli have been shown to improve cell mediated extracellular
matrix synthesis [135]. To investigate the influence of mechanical stimuli on
ovine MPC and OB, cells were cultured over 14 days under static or dynamic
conditions with osteogenic media containing 10% or 20% FBS respectively. It
was found that that the combination of fluid shear forces in dynamic culture
and 20% FBS media content significantly increased ALP activity at day 7
(followed by a significant decrease towards day 14)(Fig. 10) and deposition
of mineralized matrix at day 14 (Fig. 11, 12) for both MPC and OB cultures.
Furthermore, the media content of 20% FBS significantly stimulated ALP
activity in MPC at day 14. When comparing the osteogenic OB and MPC
groups amongst each other, the increased degree of mineralization in the
group cultured with 20% FBS on the rocking plate appeared not be an effect
of increased cell number (p-values > 0.05, Fig. 15). XPS analysis for ovine
MPC and OB cultured on Thermanox coverslips for a period of 14 days
under dynamic osteogenic conditions with 20% FBS media content showed a
characteristic double peak for calcium in MPC cultures, considerably smaller
amounts of calcium were found for OB cultures (Fig. 13, 14).
61
Fig. 10: Relative alkaline phosphatase (ALP) activity for ovine osteoblasts (OB) and
mesenchymal progenitor cells (MPC) under different culture conditions. The
combination of fluid shear forces in dynamic culture and 20% FBS media content
appeared to increase enzyme activity (n=3). Error bars represent standard
deviations.
Fig. 11: Quantification of the incorporated alizarin red s dye after 14 days showed a
significantly higher degree of mineralized extracellular matrix production for MPC
when compared to OP when subjected to a combination of fluid shear forces in
dynamic culture and 20% FBS media content (n=3). Error bars represent standard
deviations.
62
Fig. 12: The quantification of calcium incorporated in the extracellular matrix after 14
days using a Wako HR II calcium assay confirmed a significantly higher amount of
calcium for osteogenically induced MPC under dynamic culture conditions with 20%
FBS media content. The average calcium content for MPC was 0.024 mg per 24-
well compare to 0.0037 mg for OB (p<0.05)(n=3). Error bars represent standard
deviations.
63
Fig. 13: XPS analysis for native ovine bone derived from the mid diaphysis of the
tibia showing the presence of calcium with a characteristic double peak and
phosphate (A). The XPS analysis of plain Thermanox coverslips (B) as a negative
control confirmed the absence of both calcium and phosphate.
A
B
64
Fig. 14: XPS analysis for ovine MPC (A) and OB (B) cultured on Thermanox
coverslips for a period of 14 days under dynamic osteogenic conditions with 20%
FBS media content. A characteristic double peak for calcium could be detected for
MPC cultures (A), considerably smaller amounts of calcium were found for OB
cultures (B).
A
B
65
Fig. 15: PicoGreen assays for ovine MPC and OB on day 0, 1, 3, 5, and 7 after
osteogenic induction revealed that increased degrees of mineralization in the group
cultured with 20% FBS on a rocking plate (n=3) appeared not be an effect of
increased cell number but rather increased metabolic activity (p>0.05). Error bars
represent standard deviations.
3D differentiation potential of MPC and OB in vitro
Viability, morphology and osteogenic potential of ovine MPC and OB in a
three dimensional environment was assessed by FDA/PI staining, phalloidin-
DAPI staining, SEM, and micro computed tomography. After 28 days of
66
osteogenic induction under dynamic conditions, cell viability was assessed
>90% (Fig. 16 A and D). Phalloidin-DAPI staining (Fig. 16 B and E) and SEM
analysis (Fig. 16 C and F) revealed elongated, spindle-shaped, osteoblast-
like cell morphology for both OB and MPC. However, MPC seemed to have
proliferated at a higher rate on the type I collagen coated mPCL-TCP
scaffolds forming a dense, interconnected three dimensional network. Micro
CT analysis displayed mineral deposition throughout the entire thickness of
OB and MPC-constructs, compared to control constructs (Fig. 17 A-F).
Scaffolds seeded with MPC showed a significantly higher mineral volume
fraction (MVF) compared to scaffolds seeded with OB (p=0.000197) (Fig. 17
G) while no significant difference in mineral density could be found between
MPC and OB-constructs (Fig. 17 H).
Fig. 16: SEM, live-dead and phalloidin-DAPI staining (confocal laser microscopy) of
MSC and OB on mPCL-TCP scaffolds cultured for 28 days. Cell viability was
assessed >90% for both cell types (A and D). Phalloidin-DAPI staining (B and E)
and SEM analysis (C and F) revealed elongated, spindle-shaped, osteoblast-like
cell morphology for both OB and MPC forming a dense, interconnected, three
dimensional network
67
Fig. 17: !CT analysis of 3D in vitro cultures. !CT displayed mineral deposition
throughout the entire thickness of all constructs. After 28 days, medical grade PCL-
TCP scaffolds seeded with MPC (C, D) showed a significantly higher mineral
volume fraction (MVF) (G) compared to OB seeded (E, F) or cell free scaffolds (A,
B). No significant difference in mineral density was found (H)(n=5). Error bars
represent standard deviations.
Differentiation potential of MPC and OB in vivo
The developmental potential of culture-expanded, ovine MPC and OB was
assessed in vivo following transplantation into NOD/SCID mice in association
with type I collagen coated mPCL-TCP scaffolds. Transplants were
recovered after eight weeks, subjected to micro CT analysis and then
processed for histology. Micro CT analysis revealed a significantly higher
degree of ectopic bone formation for the scaffolds seeded with OB prior to
implantation (Bone volume fraction: 20.15%) when compared to MPC
(6.12%) or the respective cell free controls (0.55 %) (Fig. 18 A-C, M).
However, no significant difference could be found with regard to the mineral
density of newly formed bone matrix (Fig. 18 N).
68
Histological examination of ectopic explants, using von Kossa staining
revealed no mineralisation for the control (no cells) constructs (Fig. 18 D, G,
J) whereas both MPC (Fig. 18 E, H, K) and OB (Fig. 18 F, I, L) seeded
mPCL-TCP scaffolds formed extensive ectopic bone within the implants,
over a course of 8 weeks in vivo. Mineral nodules containing calcium, stain
black with the von Kossa staining by virtue of silver ions (positive charge)
binding with the mineralised tissue (negative portion of the calcium salt)
forming a silver salt which is black in colour. The amount of ectopic bone
formed was significantly higher for OB seeded tissue engineered constructs
compared with MPC-seeded constructs (Fig. 18 O)(p<0.05). Residual mPCL-
TCP scaffold was evident within all transplants (Fig. 18 D) evidenced by
voids in the tissue from longitudinal and transverse sectioning of the scaffold
struts. The infiltration of haematopoietic cells together with associated
adipose elements was reminiscent of native bone marrow (Fig. 18). The
formation of different tissue types within the transplanted constructs included
mineralised bone (mb), fat (f) and fibrous connective tissue (c) with clear
blood vessel (bv) formation. The predominantly formed tissue type in both
MSC and OB samples, however, was bone with mature osteocytes enclosed
in characteristic lacunae surrounded by the bone extracellular matrix.
69
Fig. 18: Histology and !CT of in vivo specimens 8 weeks post subcutaneous
transplantation into male NOD/SCID mice (n=4). !CT revealed significantly more
bone formation for OB compared with MPC (B, C) and control (A, M). No difference
in mineral density was observed (N). Combined von Kossa/van Gieson staining on
histological sections revealed extensive bone formation for MPC (E, H, K) and OB
seeded scaffolds (F, I, L), other tissue types included muscle (m) fat (f), blood
vessels (bv) and fibrous connective tissue (c). Error bars represent standard
deviations.
70
Discussion
In translational orthopaedic research, the utilisation of large preclinical animal
models is a conditio sine qua non. Animal models utilized in the field include
representations of normal fracture-healing, segmental bone defects, and
fracture non-unions [38]. The selection of a specific animal species as a
model system involves the consideration of a number of factors.
Physiological and pathophysiological analogies are essential in respect to
the scientific question under investigation. It must be manageable to operate
and observe a multiplicity of study objects over a relatively short period of
time [39-41]. Further selection criteria include costs for acquisition and care,
animal availability, acceptability to society, tolerance to captivity and ease of
housing [42]. Mature sheep and goats possess a bodyweight comparable to
adult humans and long bone dimensions enabling the use of human implants
[54]. The mechanical loading environment occurring in sheep is well
understood [55, 56]. Since no major differences in mineral composition [57]
are evident and both metabolic and bone remodelling rates are akin to
humans [58], sheep are considered a valid model for human bone turnover
and remodelling activity [59] and show comparable bone healing potential
and bone blood supply [62]. Sheep bone, however, represents human bone
physiology and anatomy much closer as animals reach an age of six to
seven years. Only at this age bone remodels from a so called plexiform bone
into bone showing secondary osteon formation, which is a hallmark of adult
human bone. Therefore, the characterisation of cell populations derived from
animals of this age is absolutely essential.
71
As described previously, bone marrow aspirates collected from adult sheep
contain a proportion of CFU-F forming cells with an incidence and
morphology highly similar to human MSC [128, 132, 136-139]. The incidence
of ovine CFU-F exceeds those described for most other animals, but lies
within the normal range reported for human bone marrow CFU-F [128, 140-
143].
To date, there is limited information available on the cell surface
characteristics of ovine bone marrow derived MPC and bone derived OB.
This can mainly be attributed to the limited availability of antibodies specific
for and cross-reacting with equivalent sheep antigens [128, 140, 144].
However, to validate sheep as a model system for research on a cellular and
molecular level, and to provide insight into fundamental processes such as
haematopoiesis, cell migration and homing, injury repair, differentiation, and
proliferation, the availability of suitable antibodies plays a key role. The cell
surface expression profile of ovine MPC and OB showed high and uniform
levels of cell surface CD29, CD44 and CD166 which have previously been
identified as markers associated with human bone marrow stromal, adipose,
and dental pulp cells [127, 130-132]. Characteristic for ovine MPC and OB
cultures was the absence of expression of the endothelial associated
adhesion marker CD31 and haematopoietic marker CD45 consistent with
expression patterns previously described for human bone marrow derived
MSC [127, 128].
When compared to ovine OB isolated from compact tibial bone, the cells
isolated from the bone marrow showed a higher proliferative potential (Fig. 3)
indicative for immature progenitor cells. Both MPC and OB followed a normal
72
growth curve reported for different human cell types consisting of a lag phase
followed by a log phase of exponential cell growth, ending with a plateau
phase in which the growth rate declined [145-148]. The steeper slope of the
MPC growth curve between day 1 and 3 after seeding results in a higher
density of cells before the rate of growth begins to decline. A high
proliferation capacity is desirable when it comes to the application of cell
based tissue engineering strategies in preclinical models since large cell
numbers generated over a relatively short period of time may be required for
various clinical applications.
After 28 days of culture under osteogenic conditions both MPC and OB were
shown to produce mineralized extracellular matrix positive for alizarin red,
osteocalcin and type I collagen. It was shown that MPC and OB can be
induced to form mineral in culture by treatment with osteogenic medium in
vitro. Osteogenic medium contains a source of phosphate, ascorbic acid, and
dexamethasone in a rich medium such as DMEM containing fetal bovine
serum (FBS) [149]. In a chelation process, alizarin red S forms complexes
with calcium and therefore allows simultaneous evaluation of mineral
distribution and inspection of fine structures by phase contrast microscopy. It
is particularly versatile in that the dye can be extracted from the stained
monolayer and readily assayed with low variability and a much wider linear
detection range than traditional calcium detection methods [150, 151].To
monitor inorganic phosphate deposition may be problematic due to the high
levels of contaminating phosphate associated with other components of the
cell and the high levels of free phosphate in the cytosol.
73
The extracellular matrix produced by ovine MPC and OB was shown to stain
positive for type I collagen and osteocalcin. Type I collagen comprises
approximately 95% of the entire collagen content of bone and about 80% of
the total proteins present in bone [152, 153]. The increase in type I collagen
expression on the gene level in MPC and its presence in the extracellular
matrix deposited by MPC under osteogenic conditions is therefore consistent
with osteogenic differentiation processes. Osteocalcin is synthesized and
secreted by normal maturing osteoblasts. It is one of the major non-
collagenous bone matrix proteins, with osteocalcin comprising 1% to 2% of
the total proteins in the skeleton [154]. Osteocalcin binds with high affinity to
hydroxyapatite crystals, the key mineral component of bone, and regulates
bone crystal formation [155]. In the MPC cultures, osteocalcin expression
was up-regulated around day 21 and further increased towards the end of
week 4 which is consistent with osteocalcin being a late osteogenic marker.
Osteocalcin protein was additionally detected immunohistochemically in the
ECM produced by both ovine MPC and OB further suggesting a bone like
composition. Osteopontin is biosynthesized by a variety of tissue types
including pre-osteoblasts, osteoblasts, osteocytes, and bone marrow cells
and is considered an early osteogenic marker and was shown to be up-
regulated in the ovine MPC cultures early during differentiation. It has further
been implicated as an important factor in bone remodelling [156]. For the
OB, no significant changes in type I collagen expression were found, only a
small increase in osteocalcin expression. Osteopontin levels increased
between day 0 and 7 and remained at this level. The findings suggest that
while MPC undergo a differentiation process when supplemented with
74
osteogenic factors, OB – as they are mature cells already – very much
maintain their gene expression pattern. Alkaline phosphatase (ALP) is
considered a marker for osteoblastic activity in vitro. ALP activity measured
at day 14 and 28 displayed a typical rise-fall pattern [133, 134] and was
significantly increased in osteogenically induced OB and MPC (p<0.05) when
compared to their respective controls (Fig. 9).
The inorganic, crystalline phase of native bone consists mainly of
hydroxyapatite, which includes calcium phosphate, calcium carbonate,
calcium fluoride, calcium hydroxide and citrate [157]. The presence of
calcium and phosphate in the matrices produced by MPC and OB as
confirmed by XPS, WAKO HRII calcium assay and alizarin red staining
suggests the deposition of a mineralized, bone-like matrix in vitro.
Culture systems that enhance mass transport and deliver controlled
mechanical stimuli have been shown to improve cell mediated extracellular
matrix synthesis in vitro [135] and it was shown that fluid shear stress can
synergistically enhance osteoblastic differentiation [158] [159]. In the present
study, for both ovine MPC and OB, it was found that the combination of fluid
shear forces in dynamic culture and 20 % FBS media content significantly
increased ALP activity at day 7 (Fig. 11) and deposition of mineralized matrix
at day 14 (Fig. 12, 13). MPC cultures seemed to be even more susceptible to
mechanical stimuli than OB. Interestingly, it was demonstrated that the
increased degree of mineralization was not a function of proliferation rather
than a result of stimulated metabolic cell activity with the higher FBS content
providing the increased demand for nutrients.
75
Since the extrapolation from results in 2D to cell behaviour in 3D is rather
difficult, 3D in vitro cultures of ovine MPC and OB were established on
medical grade PCL-TCP scaffolds produced via fused deposition modelling.
Polymer–calcium phosphate composites confer favourable mechanical and
biochemical properties for bone tissue engineering, including strength
(ceramic phase), toughness and plasticity (polymer phase), more favourable
degradation and resorption kinetics, and graded mechanical stiffness [106,
160]. Results obtained from 3D in vitro culture confirmed the findings from
2D differentiation studies with MPC demonstrating a higher proliferative and
osteogenic potential (Fig. 17, 18). Micro CT analysis revealed no significant
difference in matrix mineral density suggesting that the MPC had undergone
an osteogenic differentiation process towards osteoblast like cells actively
secreting mineralized matrix.
Preliminary analysis of the in vivo osteogenic developmental potential of
ovine MPC and OB was undertaken by subcutaneous transplantation into
immune-compromised NOD/SCID mice. Both MPC and OB demonstrated
osteogenic potential upon transplantation with type I collagen coated mPCL-
TCP composite scaffolds, as indicated by the presence of extensive deposits
of ectopic bone (Fig. 19). The observation of ectopic bone, fibrous tissue,
and haematopoiesis is analogous with studies using these scaffolds seeded
with porcine bone marrow derived progenitor cells [160].
76
Conclusion
In summary the detailed characterisation of possible cell candidates for bone
tissue engineering applications is of utmost importance. As mature sheep are
considered a valuable model for human bone turnover and remodelling,
experimental data obtained from cells derived from these animals are of
special interest. In the present study, ovine MPC isolated from bone marrow
aspirates and OB from cortical bone explants exhibited morphological,
immunophenotypical and multipotential characteristics similar to those in
humans. MPC reproducibly showed a higher osteogenic potential in vitro.
When transplanted subcutaneously, OB however displayed a higher
developmental potential in respect to bone formation. This underlines the
difficulties in extrapolating results obtained in vitro to an in vivo setting and
suggests that OB isolated from compact bone might represent a suitable
alternative cell population for tissue engineering applications.
Microenvironmental conditions in ectopic bone assays, however, again may
not be representative of specific cues transplanted cells would face in an
actual bone defect. Although the study represents an essential first step
towards the detailed characterization of ovine MPC and OB in translational
studies towards the establishment of preclinical in vivo models, further
studies are required to verify these findings in orthotopic models.
77
Chapter III
Ovine Bone and Marrow derived Progenitor Cells and their Potential for
Scaffold Based Bone Tissue Engineering Applications in vivo
78
79
Introduction
The investigation of biological processes driving tissue regeneration is a key
step in order to optimize tissue engineering approaches aiming to correct
and repair severe tissue damage necessitating complex surgical procedures
including organ and tissue transplantation. Thus, a detailed understanding of
host responses to implanted donor cells would facilitate the development of
novel clinical treatment strategies. In bone tissue engineering, some
traditional approaches rely on the transplantation of osteogenic cells seeded
onto scaffolds mimicking chemical and physical properties of native tissue
[126, 161]. Therefore, a large number of different materials has been
investigated both in vitro and in vivo to evaluate material biocompatibility as
well as their osteoinductive and osteoconductive properties [162, 163]. In a
number of bone tissue engineering studies, mesenchymal stem cells (MSC)
and osteoblasts (OB) have been used in conjunction with these materials.
Both cell types exhibit distinct intrinsic osteogenic characteristics, which
make these cells suitable candidates for bone tissue engineering
applications. Furthermore, transplantation studies have demonstrated that
MSC and OB are capable of synthesizing bone-like extracellular matrix
(ECM) [164-166]. The processes of de novo bone formation in vivo are partly
attributed to members of the transforming growth factor-" (TGF-")
superfamily, specifically the bone morphogenetic proteins (BMPs). In
particular BMP-7 was shown to have not only osteoinductive but also
angiogenic effects [167, 168]. However, the molecular and cellular events
surrounding osteoneogenesis remain poorly understood with respect to
80
neovascularisation, and the recruitment and differentiation of osteogenic
precursor cell populations.
In the present study, ovine MSC and OB were transplanted subcutaneously
into NOD/SCID mice combined with and without recombinant human BMP-7
(rhBMP-7). This was to assess the influence of donor cell commitment on
bone synthesis and to evaluate the contribution of host cells to
neovascularisation as well as formation and maturation of tissue engineered
bone. It was hypothesized that bone cell origin, ossification type, and degree
of vascularisation and bone neoformation is dependent on the nature and
commitment of transplanted cells as well as supplemented growth factors
such as rhBMP-7.
81
Materials and Methods
Isolation of ovine MSC and OB
Ovine osteoblast explants were isolated from six to seven year old Merino
sheep as approved by the animal ethics committee of the Queensland
University of Technology, Brisbane, Australia (ethics number 0700000915).
Mid-diaphyseal, compact tibial bone samples were collected, minced, and
washed prior to incubation with 10 ml 0.25% trypsin/EDTA (Invitrogen) for 3
min at 37°C, 5% CO2. After trypsin inactivation with 10 ml low glucose
DMEM containing 10% FBS (Invitrogen), samples were washed with PBS
and transferred to 175 cm2 tissue culture flasks (Nunc). Samples were
topped-up with 12 ml of DMEM containing 10% FBS and 1%
penicillin/streptomycin. Osteoblast outgrowth occurred after 5-7 days. Cells
were expanded to the second or third passage for subsequent experiments.
Bone marrow aspirates were obtained from the iliac crest under general
anaesthesia. Total bone marrow cells (0.5-1.5 x 107 cells/ml) were plated at
a density of 1-2 x 107 cells/cm2 in complete medium comprising low glucose
DMEM supplemented with 10% FBS, 100 U/ml penicillin and 100 !g/ml
streptomycin. Cells were subsequently plated at a density of 103 cells/cm2.
Scaffold fabrication and cell seeding
Bioresorbable scaffolds of medical grade )-poly-caprolactone incorporating
20% "-tricalcium phosphate (mPCL–TCP) were produced by fused
deposition modelling (FDM) as described previously (Osteopore
International, Singapore; www.osteopore.com.sg) [169]. The structural
parameters of the scaffolds were tailored by computer aided design (CAD)
82
and included 100% pore interconnectivity within a range of 350–500 µm size,
70% scaffold porosity, and a 0/90° lay down pattern. Using biopsy punches,
cylindrical scaffolds of an outer diameter of 8 mm, an inner diameter of 4 mm
and a height of 4 mm were produced. Prior to cell seeding, scaffolds were
surface treated with 1M NaOH for 6 h, washed five times with PBS, sterilized
under UV light, and transferred to 24-well plates (Nunc). Scaffolds were then
incubated with DMEM/20% FBS for 1 h before 250.000 ovine MSC or OB
suspended in 60 !l of basal medium were seeded onto each mPCL-TCP
scaffolds and placed in an incubator. After 1 h, 1 ml of medium was added to
each well. Cell scaffold constructs were cultured in DMEM/20% FBS
supplemented with 50 µg/ml ascorbate-2-phosphate, 10 mM beta-
glycerophosphate, and 0.1 µM dexamethasone on a rocking plate (f=0.125
Hz) for 4 weeks.
Cell sheet fabrication
For cell sheet fabrication, MSC and OB were seeded at a density of
3000/cm2 into 6-well plates and cultured in low glucose DMEM
supplemented with 10% FBS and 1% penicillin/streptomycin until confluent.
Cells were then cultured in osteogenic media consisting of DMEM/20% FBS
supplemented with 50 µg/ml ascorbate-2-phosphate, 10 mM beta-
glycerophosphate, and 0.1 µM dexamethasone for 4 weeks to allow
mineralization of the cell sheets. One day prior to transplantation, cell sheets
were carefully detached using a cell scraper and two sheets of matching cell
type were wrapped around each of the cell seeded scaffolds and cultured in
osteogenic media until transplantation.
83
In vivo transplantation studies
Cell scaffold constructs were transplanted subcutaneously into both left and
right side pockets formed in the dorsal surface of 21 ten week old, male,
immune compromised NOD/SCID mice (ARC, Perth, WA, Australia).
Experimental groups included
1.) mPCL-TCP scaffold
2.) scaffold + fibrin glue
3.) scaffold + fibrin glue + rhBMP-7
4.) scaffold + oMSC
5.) scaffold + oMSC + fibrin glue + rhBMP-7
6.) scaffold + oOB
7.) scaffold + oOB + fibrin glue + rhBMP-7.
An amount of 50.5 !l fibrin glue (Tisseel, Baxter) was used to administer 5
!g of rhBMP-7 (Stryker) to the inner duct of the scaffolds intraoperatively.
Each group included six implants. For experimental group 4-7 additional four
constructs with BrdU labelled cells were transplanted.
84
Fig. 1: The schematic illustrates the in vitro generation of tissue engineered
constructs (TEC) from primary cells and scaffolds as well as the method of rhBMP-7
administration at the time of transplantation.
BrdU labelling of cells
Cells for Bromodeoxyuridine (5-bromo-2-deoxyuridine, BrdU) labelling were
seeded at a density of 3000/cm2 and 250.000/scaffold in DMEM/10% FBS
and allowed to attach overnight. The day after seeding, BrdU labelling was
achieved by incubating ovine MSC or OB with the BrdU labelling reagent
(Invitrogen) at a concentration of 1:100 in DMEM/10% FBS for 5 h [170] (Fig.
2). BrdU is a synthetic nucleoside that is an analogue of thymidine. It can be
incorporated into newly synthesized DNA of replicating cells substituting for
thymidine during DNA replication thus labelling the respective cells. Specific
antibodies can then be used to visualize the incorporated chemical.
Specimens for transplantation were generated as described above,
recovered after 8 weeks, and fixed in 4% paraformaldehyde.
85
Fig. 2: Monolayer of ovine bone marrow derived cells labelled with BrdU.
Bromodeoxyuridine (5-bromo-2-deoxyuridine, BrdU) is a synthetic nucleoside that is
an analogue of thymidine. BrdU can be incorporated into newly synthesized DNA of
replicating cells substituting for thymidine during DNA replication. Antibodies
specific for BrdU can then be used to detect the incorporated chemical (brown
nuclei, arrows). The cultures were counterstained with Haematoxylin to visualize the
nuclei of non-labelled cells (asterix, violet nuclei). Magnification: 4x (A) and 10x (B).
Biomechanical testing
To determine differences in mechanical properties resulting from new bone
formation, compression testing was performed on all six specimens per
group. During the testing period, samples were kept in phosphate-buffered
saline (PBS) at ambient conditions. Tests were performed on an Instron
5848 testing system with a 500 N load cell. The specimens were
compressed at a rate of 1 mm/min up to a strain level of approximately 5%.
The compression stiffness was calculated from the stress-strain curve as the
slope of the initial linear portion of the curve.
86
!CT analysis
In vivo transplanted constructs were scanned at a voxel size of 16 !m (!CT
40, Scanco Medical AG, Brütisellen, Switzerland) and evaluated at a
threshold of 140, a filter width of 1.0 and filter support of 2. X-ray attenuation
was correlated to sample density using a standard curve generated by
scanning hydroxyapatite phantoms with known mineral density. Bone volume
fraction and mineral density were quantified throughout the entire construct.
Immunohistochemistry
For immunohistochemistry, three samples per group were fixed with 4%
paraformaldehyde, decalcified in 15% EDTA in Tris-HCl (pH 7.4) for 10 days
and embedded in paraffin [171]. Sections of 5 !m thickness were prepared,
deparaffinised and rehydrated. Subsequently, sections were rinsed in
distilled water and placed in 0.2 M Tris-HCl buffer (pH 7.4). The sections
were incubated with 2% bovine serum albumin (BSA) (Sigma, Sydney,
Australia) in DAKO antibody diluent (DAKO, Botany, Australia) in a
humidified chamber at room temperature for 20 min to block nonspecific
binding sites. Endogenous peroxidase was quenched by incubating the
sections in 3% H2O2 in Tris-HCl for 20 min. This was followed by three
washes with Tris buffer (pH 7.4) for 2 min each. Subsequently,
immunohistochemical staining was performed using a primary mouse
monoclonal antibody specific to osteocalcin (ab13418, Abcam Ltd.) and to
type II collagen (MS-235-P0, Invitrogen). Non-immunized mouse IgG
(086599, Invitrogen) was used as an isotype control to rule out non-specific
reactions of mouse IgG with ovine tissues as well as non-specific binding of
87
the secondary antibodies and/or peroxidase labelled polymer to ovine
tissues.
The sections were incubated with the specific antibody or negative control at
room temperature in humidified chambers for 30 min with the primary
antibody at a concentration of 1:200 in DAKO antibody diluent. The sections
were washed three times for 2 min with Tris buffer (pH 7.4) and incubated
with peroxidase labelled dextran polymer conjugated to goat anti-mouse and
anti-rabbit immunoglobulins (DAKO EnVision+ Dual Link System Peroxidase,
DAKO) at room temperature in humidified chambers for 30 min. Colour was
developed using a liquid 3,3-diaminobenzidine (DAB) based system (DAKO).
Mayer’s haematoxylin was used as a counterstain, and Kaiser’s glycerol
gelatin (DAKO) for coverslip mounting.
Histochemistry
For histological examination half of the specimens of each group were fixed
in 4% paraformaldehyde, and dehydrated using an ethanol gradient (30 min
in 70%, 1 h in 90%, 95% and 100% ethanol). The samples were then
processed through xylenes for 40 min three times, infiltrated with MMA for 3
h and embedded in MMA containing 3% PEG. Seven micrometre sections
were cut with an osteomicrotome (SM2500; Leica Microsystems, Wetzlar,
Germany), stretched with 70% ethanol onto a poly-lysine coated microscope
slide (Lomb Scientific), and overlayed with a plastic film. Slides were then
clamped together before being dried for 12 h at 60°C. Sections were stained
using combined von Kossa and van Gieson [129] stains to visualise the
mineralised bone and connective tissue respectively.
88
In addition, paraffin sections were stained for alcian blue/nuclear fast red and
haematoxylin/eosin using standard protocols [172, 173].
Quantification of Neovascularisation
Mallory’s trichrome (Sigma) stainings according to manufacturer’s
instructions were performed to reveal the presence of blood vessels [174].
Blood vessel formation was quantified by image analysis as described below
for mineralized tissue.
Tartrate resistant acid phosphatase (TRAP) staining
To detect osteoclast activity, TRAP staining was performed as described
previously by Erlebacher and Derynck [175]. Sections of 5-8 !m thickness
were deparaffinised in two changes of xylene, rehydrated through a graded
series of ethanols and placed in 0.2 M acetate buffer (0.2 M sodium acetate
and 50 mM L(+) tartartic acid in ddH2O, pH 5.0) and incubated for 20 min at
room temperature. Sections were then incubated with 0.5 mg/ml naphtol AS-
MX phosphate and 1.1 mg/ml fast red TR salt (both Sigma) in 0.2 M acetate
buffer for 1-4 h at 37°C until osteoclasts appeared bright read. Sections were
then counterstained with haematoxylin, air-dried and mounted.
Detection of BrdU-labelled cells
BrdU labelled specimens were fixed with 4% paraformaldehyde decalcified in
EDTA for 10 days and embedded in paraffin using an automated tissue
processor (Excelsior ES, Thermo Scientific). Sections of 3-4 !m thickness
were prepared, deparaffinised and rehydrated. For detection of BrdU labelled
89
cells, a Zymed® streptavidin-biotin based system for BrdU staining
(Invitrogen) was used according to the manufacturer’s protocol.
Quantification of BrdU positive cells located within newly formed bone was
achieved as described in detail by Tatapudy et al. [176].
SEM and EDX
To confirm the presence of calcium and phosphate in areas histologically
identified as bone, scanning electron microscopy (SEM) in combination with
energy dispersive X-ray spectroscopy (EDX) was performed on 7 !m
sections of the PMMA embedded samples on a JEOL JSM 6300 Scanning
Electron Microscope (JEOL, Tokyo, Japan) [177, 178].
Image analysis
Histology sections were quantified using Image J software to determine the
amount of mineralisation in a given area of a section. Briefly, a JPEG image
of the entire tissue section stained for von Kossa/van Gieson was selected,
converted to greyscale, and a scale bar was calibrated onto the image. The
entire tissue section area was then calculated by segmenting the entire
tissue region from the background, and then measuring the area. Next, only
the mineralised (black) area was segmented from the entire tissue area, and
measured. The total mineralised area was then calculated as a percentage
of the total section area. Six sections were analysed per sample group.
To quantify the level of blood vessel formation low-magnification micrographs
of paraffin sections stained for Mallory’s trichrome were captured, and the
area covered by blood vessels was calculated using Image J and expressed
90
as a percentage of the total tissue area [179]. Six sections were analysed per
sample group.
Statistical analysis
TStatistical analysis was carried out using the student’s t test and p values <
0.05 were considered significant (SPSS, SPSS Inc., Chicago, Illinois, USA).
Experiments were repeated three times, results are presented as mean
values +/- standard deviation.
91
Results
Biomechanical testing
When compared to the cell free control groups or constructs including MSC,
compression testing at ambient conditions showed higher stiffness values for
mPCL-TCP scaffolds seeded with OB (58 N/mm, SD=3.52) and a significant
increase in compressive stiffness when rhBMP-7 was administered
intraoperatively to OB seeded scaffolds (71 N/mm, SD=5.8)(Fig. 3).
Fig. 3: Biomechanical testing was performed and compressive stiffness values
determined. The transplantation of osteoblasts (OB) or osteoblasts with rhBMP-7
(PCL-OB-BMP) led to increased stiffness when compared to controls or groups
including MSC (n=6).
!CT analysis
MicroCT analysis was performed to determine the bone volume fraction and
bone mineral density throughout the transplanted constructs. 3D
reconstructions of scanned specimens (Fig. 4) showed no bone formation in
the scaffold only (Fig. 4 A) or scaffold-fibrin (Fig. 4 B) group and only
scattered bone formation in scaffolds seeded with MSC (Fig. 4 C). More
92
deposited bone was observed when scaffolds or MSC seeded scaffolds were
combined with rhBMP-7 (Fig. 4 D, E). However, the highest amount of bone
deposition was observed in OB seeded scaffolds (Fig. 4 F) and especially
when the OB seeded scaffolds were additionally stimulated with rhBMP-7
(Fig. 4 G). The visual, macroscopic impression based on the 3D
reconstructions was confirmed by quantitative analysis of the bone volume
fraction within the transplanted specimens. In the mPCL-OB-BMP group the
proportion of ectopic bone was found to be 17% (SD=3.04) followed by 12%
(SD=2.8) in the PCL-OB group and 2-5% in all other groups (Fig. 5 A). No
significant difference in bone mineral density was found between the different
groups (Fig. 5 B).
Fig. 4: 3D !CT reconstructions 8 weeks after subcutaneous transplantation into
NOD/SCID mice. Transplantation of the mPCL scaffold only (A) and scaffold with
fibrin glue (B) did not lead to ectopic bone formation. Only little bone formation was
found when the scaffold was seeded with MSC (C) and combined with rhBMP-7 (D).
The scaffold plus rhBMP-7 group showed slightly more bone formation (E) whilst the
highest amounts were seen in the groups including OB (F) or OB with rhBMP-7 (G).
93
Fig. 5: Quantitative analysis of bone volume fraction (A) and bone mineral density
(B) 8 weeks after transplantation. The highest amounts of bone formation were
observed for scaffolds seeded with OB or OB seeded scaffolds combined with
rhBMP-7 (PCL-Fibrin-OB-BMP). No significant difference in bone mineral density
between the different groups was detected (n=6). The asterix indicates statistical
significance. Error bars represent standard deviations.
Histology
Eight weeks after transplantation into NOD/SCID mice, Haematoxylin and
Eosin (H&E) staining of paraffin sections revealed that residual mPCL-TCP
scaffold was evident within all transplants (Fig. 6 A-P, asterix) evidenced by
A
B
94
voids within the tissue from longitudinal and transverse sectioning of the
scaffold struts. The formation of different tissue types within the transplanted
constructs included mineralised bone (NB), adipose tissue (AT) and fibrous
connective tissue (CT) with clear blood vessel (BV) formation. The
predominant tissue type in the scaffold only group represented fibrous
connective tissue (Fig. 6 A-D) with scattered adipose islets. Similar findings
were observed in the scaffold-fibrin group (Fig. 6 E-H) although a higher
proportion of adipose tissue was evident. In the scaffold-BMP group (Fig. 6 I-
L) islets of newly formed bone, mainly located at the periphery of the
scaffold, intermingled with the surrounding fibrous connective and adipose
tissue. Both scaffold-MSC and scaffold-MSC-rhBMP-7 group (Fig. 7 A-D and
I-L) showed primarily peripheral bone formation along the outer boundaries
of the scaffold also lining the inner duct while a higher degree of bone
formation was observed in the scaffold-MSC-BMP group. The main tissue
type in central regions of the scaffold was adipose with smaller proportions of
connective tissue.
The predominantly formed tissue type in both groups including transplanted
OB, however, was bone with mature osteocytes (arrows) embedded in
characteristic lacunae surrounded by the bone extracellular matrix. The bone
tissue was evenly distributed in central and peripheral regions of the scaffold.
However, larger amounts of newly formed bone could be observed when
osteoblasts were transplanted in combination with rhBMP-7.
Qualitative assessment of histological sections from each of the groups
showed a higher degree of neovascularisation with blood vessels of larger
diameters within all transplanted constructs including rhBMP-7.
95
A combined von Kossa-van Gieson stain was performed on 8 !m sections of
PMMA embedded specimens to specifically visualise and quantify the
mineralised bone within the tissue engineered constructs. The stain principle
is based on a photochemical reaction in which silver ions react with
phosphate. The staining confirmed the results obtained by !CT analysis with
the significantly highest amount of mineralisation found in the PCL-OB-BMP
group (Fig. 8, 9 and 11). Interestingly, bone lining cells in the groups
including MSC were of round, hypertrophic appearance and reminiscent of
chondrocytes. Bone synthesizing cells in the two groups including OB,
however, were elongated, spindle-shaped and of fibroblast-like morphology
(Fig. 12).
Fig. 6: H&E staining on histological paraffin sections 8 weeks after transplantation
revealed no bone formation in the scaffold only (A-D) and scaffold-fibrin group (E-H)
and only a small amount, peripheral ectopic bone formation (NB) for the scaffold-
BMP group (I-L). Other tissue types included fat (AT), blood vessels (BV) and
fibrous connective tissue (CT). Residual mPCL-TCP (&) evidenced by voids within
the tissue. Scale bars: B, F, J: 200 !m; C, G, K: 100 !m; D, H, L: 50 !m.
96
Fig. 7: After 8 weeks, H&E staining on paraffin sections revealed little, bone
formation in the scaffold-MSC group with large proportions of adipose tissue (AT)
(A-D). More ectopic, evenly distributed, new bone (NB) was observed in the
scaffold-OB group (E-H). With both MSC (I-L) and OB (M-P) the addition of rhBMP-
7 lead to increased deposition of mineralised bone. However, larger amounts of
bone with mature osteocytes (arrows) enclosed in characteristic lacunae were
observed in the scaffold-OB-BMP group while lower proportions of adipose (AT) and
connective tissue (CT) were evident. Residual mPCL-TCP (&) resulted in voids
within the tissue. Scale bars: B, F, J, N: 200 !m; C, G, K, O: 100 !m; D, H, L, P: 50
!m.
The vascularisation of an implant is one of the first responses of the host to a
graft and is of utmost importance for its survival and functionality. Therefore,
the blood vessel formation was quantified on histological sections. Image
analysis showed a higher amount of neovascularisation in all groups that
included the administration of rhBMP-7 (Fig. 10) independent of transplanted
cell type.
97
Fig. 8: Sections of PMMA embedded cell free control specimens stained for von
Kossa/van Gieson. No bone formation was observed in the PCL and PCL-fibrin
group, only little in the PCL-BMP group. Besides bone (NB, black), identified tissue
types included muscle (M), connective tissue (CT), adipose tissue (AT), and blood
vessels (BV). Scale bars: B, F, J: 200 !m; C, G, K: 100 !m; D, H, L: 50 !m.
Fig. 9: Bar graph demonstrating the results of the histomorphometric analysis of
PMMA embedded sections stained for von Kossa-van Gieson. No mineralized
tissue was found in the scaffold only group (PCL, 1%) and scaffold-fibrin group
(PCL-Fibrin, 1%). No significant difference in bone formation was evident between
the scaffold-BMP (PCL-BMP, 17%), scaffold-MSC (PCL-MSC, 19) and scaffold-
MSC-BMP (PCL-MSC-BMP, 26%) groups. Osteoblasts transplanted in combination
with the PCL-scaffold (PCL-OB) formed significantly more ectopic bone (47%). The
addition of BMP to scaffold-OB constructs caused an even further increase in
deposition of mineralised extracellular matrix (PCL-OB-BMP, 67%). Error bars
represent standard deviations (n=6).
98
Fig. 10: The figure illustrates the amount of neovascularisation within the constructs
8 weeks after transplantation. Blood vessels are expressed as a percentage of total
tissue area. The administration of rhBMP-7 appeared to stimulate blood vessel
formation. Error bars represent standard deviations.
Fig. 11: After 8 weeks, von Kossa-van Gieson staining on 8 !m PMMA sections
revealed little bone formation in the scaffold-MSC group and large amounts of
adipose tissue (AT) (A-D). More new ectopic bone was observed in the scaffold-OB
group (E-H). With both MSC (I-L) and OB (M-P) the addition of rhBMP-7 lead to
increased deposition of mineralised bone (NB). However, larger amounts of bone
were observed in the scaffold-OB-BMP group. On the other hand lower proportions
of adipose (AT) and connective tissue (CT) were evident. Residual mPCL-TCP (&)
resulted in voids within the tissue. Scale bars: B, F, J, N: 200 !m; C, G, K, O: 100
!m; D, H, L, P: 50 !m.
99
Fig. 12: High magnification images of bone lining cells. The bone forming cells in
both groups including transplanted MSC (A, C) appear of round, hypertrophic
morphology (arrows) with big nuclei. Both in the scaffold-OB (B) and scaffold-OB-
BMP group (D), the cells immediately lining the ectopically formed bone (black)
were of elongated, spindle-like shape (arrows). Scale bars: 20 !m.
Osteocalcin, an extracellular matrix protein, is partly incorporated into the
bone matrix, partly delivered to the circulatory system. It is considered a late
and specific osteogenic marker determining terminal osteoblast
differentiation regulating bone crystal formation. Osteocalcin (OC) expression
was examined in all experimental groups in which new bone formation was
evident (Fig. 13). In all groups, strong extracellular, matrix associated and
cytoplasmatic OC expression was detected in osteocytes enclosed in the
100
mature bone (arrow heads). Notably, bone lining osteoblasts (arrows) also
expressed high levels of OC.
Fig. 13: Immunohistochemical staining for osteocalcin (OC) within newly formed
bone (NB) counterstained with haematoxylin demonstrating high extracellular
and cytoplasmatic osteocalcin expression in both osteocytes enclosed in
extracellular matrix (arrow heads) and bone lining osteoblasts (arrows). Fibrous
connective tissue (CT) and adipose tissue (AT) did not stain positively for OC.
Only light, haematoxylin caused background staining is evident in the isotype
controls (C, F, I, L, O). Scale bars: left and right column 100 !m, middle column
50 !m.
101
In general, direct and indirect (chondral) ossification processes can be
distinguished. Chondral ossification requires preformed cartilaginous tissue
which is subsequently replaced by bone. This cartilaginous tissue contains
high amounts of glycosaminoglycans. Alcian blue is a phthalocyanine dye
which stains acid mucopolysaccharides and glycosaminoglycans (GAGs)
that appear blue to bluish-green. Therefore, alcian blue staining was
performed on all groups in which bone formation was evident to evaluate
whether direct or indirect ossification had occurred in the transplanted
specimens (Fig. 14). In all groups GAGs could be detected (arrows) within
areas of newly deposited bone. While only little GAG remnants were evident
within the bony extracellular matrix in the scaffold-OB (Fig. 14 E and F) and
scaffold-OB-BMP (Fig. 14 I and J) group, larger amounts of GAGs were
found in the scaffold-MSC (Fig. 14 C and D) and especially scaffold-MSC-
BMP group (Fig. 14 G and H).
Type II collagen is considered a cartilage specific extracellular matrix protein.
To further confirm the presence of cartilaginous tissue in the transplanted
specimens, immunohistochemical staining was performed. Positive staining
could only be detected in the scaffold-MSC and scaffold MSC-BMP group
(Fig. 14, C and G).
102
Fig. 14: Alcian blue staining of paraffin sections counterstained with nuclear fast
red 8 weeks after subcutaneous transplantation. The images show evidence of
glycosaminoglycan (GAG, blue) remnants (arrows) within the newly formed
bone (NB) and adjacent connective tissue (CT) in the scaffold-BMP (A, B),
scaffold-OB (E, F), and Scaffold-OB-BMP (I, J). Larger GAG deposits were
found within the extracellular matrix in the scaffold-MSC and scaffold-MSC-BMP
group (C, D and G, H). Large proportions of adipose tissue (AT) were found in
groups including MSC. Residual mPCL-TCP was present in all groups (&) as
evidenced by voids within the tissue. Scale bars: left column 100 !m, right
column 20 !m.
103
Fig. 15: Immunohistochemistry for type II collagen. The images demonstrate
positive staining results (arrows) for the Scaffold-MSC (C) and Scaffold-MSC-
BMP (G) group within the connective tissue (CT) lining the newly formed bone
(NB). No collagen II deposition was evident in all other groups. Scale bars: 100
!m.
Tartrate resistant acid phosphatase (TRAP) is a glycosylated monomeric
metalloenzyme expressed in mammals. Under physiological conditions,
TRAP is highly expressed by osteoclasts, activated macrophages, and
neurons. In osteoclasts, TRAP is commonly localized within the ruffled
border area, within lysosomes, and in Golgi cisternae and vesicles. TRAP
104
positive cells can be stained using a colorimetric reaction resulting in
bright red colour deposition. TRAP staining on paraffin embedded
specimens showed high osteoclastic activity in all groups including
rhBMP-7 (Fig. 16 G-O) while no colour deposition was observed in the
other experimental groups (Fig. 16 A-F). In case of positive staining,
osteoclasts were lining the outer boundary of areas of new bone
formation in close proximity to the surrounding fibrous connective tissue
(Fig. 16 G-O, arrows).
Bromodeoxyuridine (5-bromo-2-deoxyuridine, BrdU) is a synthetic
nucleoside that is an analogue of thymidine. BrdU is commonly used in
the detection of proliferating cells in living tissues. BrdU can be
incorporated into the newly synthesized DNA of replicating cells during
the synthetic phase of the cell cycle, substituting for thymidine during
DNA replication. Antibodies specific for BrdU can then be used to detect
the incorporated chemical, thus indicating cells that were actively
replicating their DNA. In any case where labelled cells were transplanted,
immunohistochemistry showed positive staining for BrdU in cells
embedded in newly formed bone matrix as well as in bone lining cells
clearly indicating that the transplanted cells had actively contributed to
neoosteogenesis. However, cells positive for BrdU were also detected in
bone surrounding adipose and connective tissue. The amount of labelled
cells within bone was higher in groups including rhBMP-7, in tissues
other than bone it appeared to be higher in the groups including MSC
transplantation suggesting a higher degree of plasticity of the less
105
committed MSC when compared to compact bone derived, differentiated
OB (Fig. 17).
Fig. 16: TRAP staining of paraffin-embedded specimen counterstained with
haematoxylin. In all groups including rh-BMP-7, positive, bright red staining was
clearly evident (arrows) in areas between newly formed bone (NB) and
connective tissue (CT). Residual mPCL-TCP (&) was evident represented by
voids within the tissue. Scale bars: left column 200 !m, middle column 100 !m,
right column 50 !m.
106
Fig. 17: BrdU staining of paraffin-embedded specimen counterstained with
haematoxylin. Cells positive for BrdU were found within the newly formed bone (NB)
(arrows), as bone lining cells (arrow heads), and within fat (AT) and connective
tissue (CT). The amount of BrdU positive cells was significantly higher in groups
including rhBMP-7. Residual mPCL-TCP (&) was evident. Scale bars: 50 !m.
I
107
SEM and EDX of areas identified as bone (Fig. 18 *) on histological PMMA
sections revealed an identical spectrum of composing elements for both OB
(Fig. 18 A and E) and MSC (Fig. 18 B and F) derived bone with high
amounts of calcium (Fig. 18, Ca, arrow heads) and phosphate (Fig. 18, P,
arrows). No calcium or phosphate could be detected in bone lining soft tissue
(Fig. 18, asterisks).
Fig. 18: Backscattered electron microscopy images on PMMA sections of OB (A)
and MSC derived (B) bone (*) and surrounding soft tissue (&). The EDX spectra of
OB and MSC derived bone (E and F) showed and identical elementary composition
with high amounts of calcium (arrow heads) and phosphate (arrows).
108
Discussion
Mature bone consists of specialized cells secreting and remodelling the
surrounding extracellular matrix (ECM). Osteoblasts (OB), which
subsequently mature into osteocytes, synthesize and deposit the
proteinaceous ECM, which then becomes calcified, and secrete growth
factors such as hepatocyte growth factor (HGF), fibroblast growth factor
(FGF), and bone morphogenetic proteins (BMPs) stimulating proliferation
and osteogenesis. Osteoclasts are derived from the monocyte-macrophage
lineage and play a major role in bone remodelling. The ECM is composed of
collagenous proteins (predominantly collagen type I), non-collagenous
proteins (e. g. osteocalcin, osteopontin, osteonectin and bone sialoprotein)
and mineralized matrix (hydroxyapatite).
The use of autologous cells from the individuals’ tissue has shown promise in
the field of tissue engineering. This approach however necessitates the
removal of tissue from the individual, and in vitro isolation and expansion
before re-implantation to the site of intervention. Although, from an
immunological stance, the replacement of tissue with autologous cells is an
ideal situation, there are associated problems. Firstly, the harvest of tissue to
allow cell isolation and expansion requires surgical intervention which would
be on par with the grafting processes described previously. In addition, some
cell populations from primary sources have a low propensity for division thus
the expansion of such cells may prove problematic. This is sometimes
compounded with cellular senescence – a phenomenon describing the cease
of cellular division in primary cells, usually caused by a shortening of
109
telomere length [180]. These limitations have made the search for cell
populations, which can be expanded in culture before implantation, of the
utmost importance.
OB can be isolated from mature cancellous or compact bone and expanded
in vitro. OB originate from mesenchymal cells of the marrow stroma. These
stromal cells provide all of the benefits of primary cells, however, have the
ability to undergo multi-lineage differentiation and a higher propensity for cell
division. These cells can easily be isolated from bone marrow aspirates and
have been shown to be able to differentiate into the osteogenic, myogenic,
chondrogenic and neurogenic lineages. Therefore they have been proposed
as the most suitable cell type for bone engineering applications, a direct
comparison between osteoblasts and MSC in regards to their regenerative
potential of bone in vivo has never been made.
In the present study, the osteogenic potential of in vitro expanded ovine OB
and MSC was therefore investigated in a murine model of ectopic bone
formation. It was found that OB displayed a higher rate of ossification when
compared to MSC as quantified by histomorphometry, !CT analysis, and
compression testing. These results are consistent with previous studies (see
chapter II)[174]. EDX analysis and immunohistochemistry for osteocalcin
confirmed that the mineralized extracellular matrix synthesized by both MSC
and OB resembled mature bone. Osteocalcin is a bone-specific protein of
46–50 residues that undergoes post-translational modification by vitamin-K-
dependent +-carboxylation of three glutamic acid residues. Osteocalcin is
110
expressed by mature osteoblasts, binds strongly to hydroxyapatite, is stored
in bone matrix and released into the circulation [181, 182]. It is considered a
late and specific osteogenic marker determining terminal osteoblast
differentiation regulating bone crystal formation [155]. Histologically, the bone
synthesized by OB appeared to be of higher maturity as evidenced by early
osteoclast mediated remodelling and secondary organization with
ossification starting in close apposition to the scaffold struts.
The histological and immunohistochemical examination showed that both the
transplanted MSC and OB were viable, had mitotic activity and have
contributed to new bone formation by secreting bone ECM. BrdU labelling
seemed to have no negative effect on the amount of bone formation
observed.
In all specimens including MSC, a significantly higher proportion of adipose
tissue and less bone was observed. This underlines the difficulties in
extrapolating results obtained in vitro, where MSC reproducibly exhibited a
higher osteogenic potential (see chapter II), to an in vivo setting. The findings
suggest that OB isolated from compact bone might represent a suitable
alternative cell population for tissue engineering applications. It must,
however, be taken into account that microenvironmental conditions in ectopic
models of bone formation may again not be representative of specific cues
transplanted cells experience in orthotopic sites. Less committed cells of
high plasticity such as MSC might - once transplanted subcutaneously -
respond to a higher degree to growth and differentiation/dedifferentiation
initiating factors released by the surrounding adipose, fibrous and muscular
tissue. This, in conjunction with missing mechanical stimuli experienced in an
111
orthotopic bony environment, might prevent MSC from differentiating into
bone synthesizing osteoblasts.
As mentioned before, bone morphogenetic proteins (BMPs) are members of
the TGF-" superfamily. Of the BMPs that have displayed clinical utility for de
novo bone formation in vivo, BMP-2 and BMP-7 (osteogenic protein 1, OP-1)
are commercially available through the use of recombinant DNA technology.
In particular BMP-7 was shown to have osteoinductive effects in a variety of
studies [167, 183-187]. When transplanted in combination with ovine OB,
rhBMP-7 lead to a significant increase in bone formation while only little
effect was observed with MSC. One reason might be the dose dependent
response of MSC to BMPs both in vitro and in vivo [188-190]. The
administered dose of 5 !g might not have been sufficient to outplay the
continuous release of tissue specific growth factors from the surrounding
muscle or fat and to stimulate osteogenic MSC differentiation in an ectopic
model while on the other hand the BMP mediated signal was adequate to
excite a response in the more committed OB.
Tartrate-resistant acid phosphatase (TRAP) is a glycosylated monomeric
metalloenzyme expressed in mammals. It has a molecular weight of
approximately 35 kDa, a basic isoelectric point (7.6-9.5), and optimal activity
in acidic conditions. TRAP is synthesized as latent proenzyme and activated
by proteolytic cleavage and reduction. Under physiological conditions, TRAP
is highly expressed by osteoclasts, activated macrophages, and neurons
[191-193]. The exact physiological role(s) of TRAP is unknown, but many
functions have been attributed to this protein. In knockout mice studies,
those with a phenotype of TRAP-/- showed mild osteopetrosis, with greatly
112
reduced osteoclast activity, resulting in thickening and shortening of the
cortices, the formation of club-like deformities in the distal femur, and
widened epiphyseal growth plates with delayed mineralization of cartilage, all
of which increased with age. Likewise in TRAP overexpressing transgenic
mice, mild osteoporosis occurred along with increased osteoblast activity and
bone synthesis. Proposed functions of TRAP include osteopontin/bone
sialoprotein dephosphorylation, the generation of reactive oxygen species,
iron transport, and as a cell growth and differentiation factor [194, 195].
Amongst others, osteoclast and TRAP activity have been shown to be
stimulated by rhBMP-7 [187, 196, 197] which might have resulted in positive
staining in all groups including the administration of 5 !g rhBMP-7 while no
positive staining for TRAP was observed in all other experimental groups.
If BMPs come in contact with osteoblasts they stimulate the production of
angiogenic factors, such as vascular endothelial growth factor and fibroblast
growth factor [198]. This then leads to the recruitment and activation of
endothelial cells necessary for new blood vessel formation. It was also
demonstrated that BMPs have direct effects on endothelial cells stimulating
migration, proliferation and tube formation. Mice with knockout phenotypes of
different BMPs often have cardiovascular problems, indicating an important
role for BMPs in angiogenesis [199, 200]. These angiogenetic effects of
BMPs in the literature might have contributed to an increased amount of
neovascularisation in transplanted specimens including rhBMP-7 as
observed in the present study.
Development and formation of the skeleton (ossification) occurs by two
distinct processes - intramembraneous and endochondral ossification. Both
113
intramembraneous and endochondral ossification occur in close proximity to
vascular ingrowth. Intramembraneous ossification is characterized by
invasion of capillaries into the mesenchymal zone and the emergence and
differentiation of mesenchymal cells into mature osteoblasts. These
osteoblasts constitutively deposit bone matrix, leading to the formation of
bone spicules. These spicules grow and develop, eventually fusing with other
spicules to form trabeculae. As the trabeculae increase in size and number
they become interconnected, forming woven bone, a disorganized weak
structure with a high proportion of osteocytes, which eventually is replaced
by more organized, stronger lamellar bone. Intramembraneous ossification
occurs during embryonic development and is involved in the development of
flat bones in the cranium, various facial bones, parts of the mandible and
clavicle and the addition of new bone to the shafts of most other bones. In
contrast, bones of load-bearing joints form by endochondral formation.
Chondral ossification requires preformed cartilaginous tissue which is
subsequently replaced by bone.
In the present study, bone and especially its surrounding tissue formed after
transplantation of MSC stained positive for type II collagen and alcian blue
confirming the presence of high amounts of glycosaminoglycans
characteristic of cartilage. The findings suggest that MSC mediated bone
formation occurred by endochondral ossification through hypertrophic,
chondrocyte-like cells confirming results by other research groups [174, 201].
Furthermore, bone synthesized by OB stained only weakly for GAG and
negative for type II collagen while the adjacent connective tissue was entirely
negative for both GAG and collagen. This could indicate that the
114
transplantation of OB leads to intramembraneous ossification by elongated,
spindle-shaped osteoblasts [174]. The small amounts of GAG found within
the bone synthesized by OB might be attributed to chondroitine-4-sulfate and
hyaluronic acid both of which are components of bony extracellular matrix
[202]. Nevertheless, the low GAG amounts could also indicate that bone
formation by transplanted OB follows endochondral mechanisms and may
have progressed to an advanced stage of bone maturation.
Summary
In the present study, ovine MSC isolated from bone marrow aspirates and
OB from cortical bone explants were compared regarding their osteogenic
potential after transplantation in vivo. It was found that transplanted cells
show a high degree of survival and actively contributed to enchondral
osteogenesis. When compared to MSC, OB showed a higher degree of bone
deposition while OB derived bone was of higher maturation. Additional
stimulation with rhBMP-7 increased the rate of bone synthesis for both MSC
and OB but also increased neovascularisation and osteoclast activity. These
results suggest that origin and commitment of transplanted cells highly
influence the type and degree of ossification; furthermore that rhBMP-7
represents a powerful adjuvant for bone tissue engineering applications and
that mature bone might be an adequate alternative cell source in the context
of bone regeneration.
115
Chapter IV
Establishment of a Preclinical Ovine Model for Tibial Segmental Bone
Defect Repair by Bone Tissue Engineering Methods
116
117
Introduction
The concepts of tissue engineering have matured over the last two decades
and, when reviewing the field, it can be concluded, that the translation of
tissue engineering strategies from bench to bedside is a difficult, expensive
and time-consuming process. Hollister recently addressed the different
issues pertaining to the translation and commercialization of tissue-
engineered constructs (TEC) [203]. He argues that the so-called “Valley of
Death”, which describes the lag between research and commercialisation, is
highly prevalent in the area of tissue engineering/regenerative medicine and
originates from the high costs associated with technology development and
the need for funding preclinical animal models and clinical trials for
regulatory approval.
A set of harmonized standards (ISO 10993) that address requirements for
the biological evaluation of medical devices has been developed by the
International Organization for Standardization (ISO). Compliance with ISO
standards is required throughout Europe. The US food and drug
administration (FDA) issued the blue book memorandum #G95-1 that
adopted ISO nomenclature for device categories and included an FDA-
modified flowchart designating the type of testing needed for each device
category, and made several modifications to the testing requirements
outlined in ISO 10993-1, adding various requirements in several device
categories. Briefly, device manufacturers were asked to determine which
kinds of biological effects are of concern for the materials in a particular
device, based on the nature and duration of the product's end use. Such
effects can include sensitization, irritation, haemocompatibility, various other
118
types of toxicity, and reproductive or developmental changes. In all, 12
categories of potential biological concerns are identified in both the FDA
memorandum and ISO 10993 to conclusively confirm the safety and efficacy
of the TEC under investigation.
The lack of translation in the field of bone engineering might be related to
difficulties in integrating individual technical discoveries into model tissue
engineering systems, in manufacturing scale up, in funding, and in
regulatory approval. To establish a tissue engineering concept in a clinical
setting, a rigorous demonstration of the level of therapeutic benefit in
clinically relevant animal models is a conditio sine qua non. To closely
simulate human in vivo conditions, and to assess the effects of implanted
bone grafts and tissue engineered constructs on segmental long bone
defect regeneration, a number of large animal models have been
developed. In a recent literature review [126], however, it was concluded
that many of the preclinical models published are not well described, defined
and standardized and therefore difficult to reproduce. Only a small segment
of the research community works on a paradigm shift to translate tissue
engineering of bone into the clinical orthopaedic and reconstructive surgery
area. Therefore, in this thesis, effort was concentrated on the establishment
of a well characterized, standardized preclinical animal model (Fig. 1). In the
present chapter, the approach is illustrated of how to translate tissue
engineering of bone from bench to bedside by presenting a road map which
describes the validation of the functionality of a model to study the
regeneration of critical-sized segmental bone defects in a highly load-
bearing large animal.
119
Fig. 1: Road map to establish a critical sized bone defect model in a large animal.
Regulatory framework
Within the European Union (EU), the task of harmonizing and regulating
medical devices is handled by the European Commission in close
cooperation with member state health authorities. The purpose of the EU
harmonization effort is to merge the differing national requirements into one
law, which can be applied throughout the European Union. Legislation
adopted through this process covers implantable, non-implantable, and in
vitro diagnostic medical devices to provide manufacturers with the basis to
certify their compliance with EU-wide safety requirements. The task of
complying with essential requirements can be simplified by voluntarily using
EU harmonized standards.
120
The Commission has listed over a hundred EU wide harmonized medical
device standards addressing various essential requirements. These
standards have been developed and/or identified by the European
standards organizations (CEN, CENELEC and ETSI) and published in the
Official Journal (the EU equivalent of the U.S. Federal Register), thereby
giving them special recognition as so-called harmonized standards linked to
EU legislation. All other existing standards not published in the Official
Journal are either national or industry standards. The European harmonized
standards for medical devices are often based on international standards
(ISO or IEC). For example, the European committee for standardization
(CEN) lists a number of mandated standardization projects for the biological
evaluation of medical devices (http://www.cen.eu/cenorm/homepage.htm).
In the EU, a company that intends to commercialize a medical product may
submit a single application to the European Medicines Agency (EMEA) for a
'marketing authorisation' that is valid simultaneously in all EU Member
States, plus Iceland, Liechtenstein and Norway. This procedure is referred
to as the 'centralised authorisation procedure'. The decisions about the
authorisation of medical products are based on an objective, scientific
assessment of their quality, safety and efficacy. Conducting these
assessments within the EU is the primary role of the EMEA. Through its
scientific committees, the EMEA assesses every medicine for which a
marketing-authorisation application has been submitted, and prepares a
recommendation that is then relayed to the European Commission, which
has the ultimate responsibility for taking decisions on granting, refusing,
121
revoking or suspending marketing authorisations. The Food and Drug
Administration (FDA) is a scientific, regulatory, and public health agency
whose authority includes overseeing the marketing of products relevant to
medical practice in the United States. Devices are classified based on the
extent of oversight needed to ensure public safety. Divisions within the FDA
provide specific expertise regarding drugs, devices, biologic products and
combinations thereof. Various pathways exist to apply for marketing through
the FDA, depending on the nature of the product and its intended use.
Expert panels advise the agency on issues related to product safety and
efficacy. The development of a medical device from concept to product-
launch typically takes between 4-10 years and costs between 5 and 300
million dollars depending on the complexity of the device, its classification,
and required regulatory process [204]. The FDA recommends provision of
detailed information on the chemical composition and physical properties of
the device under review. Device performance tests include bench testing
(pH testing, dissolution and stability), biocompatibility testing and evaluation
in an animal model. According to the “Class II Special Controls Guidance
Document: “Resorbable Calcium Salt Bone Void Filler Device” issued in
2003 (http://www.fda.gov/cdrh/ode/guidance/855), the respective animal
model should
- Be representative of the indications for use
- Cover the full range of anatomical sites proposed for use, and
specifically how the device is to be used
- Include the use of skeletally mature animals and a critical size defect
- Include use of the predicate device and/or autogenous bone graft as
122
the positive control group(s) and use of empty defect as a negative
control
- Involve the use of radiography, histology, and histomorphometry to
assess bone formation and device resorption at various, relevant time
points over the course of healing, in addition to supportive
biomechanical testing of the new bone formed
- Aim at adequate study duration to demonstrate bone healing and the
effects of any residual device material.
In order to translate tissue engineering based therapies from bench to
bedside, it is mandatory to rigorously demonstrate both the level of
therapeutic benefit in clinically relevant animal models and sufficient
understanding of the mechanisms by which scaffold/cell or scaffold/growth
factor constructs integrate into host sites, form new bone and restore
function. To achieve this aim in this thesis, it was decided to validate the
functionality of engineered bone constructs of clinically relevant size in a
suitably highly load-bearing large animal model such as sheep. The use of
large animals lies in the high value of the data obtained from these models
pertaining to
(i) scaffold biomechanical properties
(ii) scaffold degradation & resorbability; and
(iii) survival of transplanted cells or release kinetics of growth factors
(which are all critical parameters to the ultimate efficiency of bone
constructs).
123
The high value of the data is attributed to great similarities to human
conditions in regards to bone biology, mineral composition, turnover and
remodelling and the mechanical loading environment.
Bone defect fixation can generally be achieved by plate fixation,
intramedullary nailing, or the application of an external fixator (Table 1). In
humans, intramedullary nailing is a commonly chosen treatment modality for
diaphyseal fractures of the lower extremities. Since these central load
carriers are less susceptible for tilting in the frontal plane, custom made
intramedullary nailing systems have been applied for fixation of large
segmental bone defects in animal models [4, 93, 97, 98]. In the context of
tissue engineering, however, intramedullary stabilization can impede the
placement of a solid, one-piece load-bearing scaffold and necessitates two
surgical approaches. Alternatively, external fixators have been widely used
as they offer versatility and ease of application [3, 12, 96]. However, with
external fixators, healing periods are reported to be significantly longer
[205]. External fixators represent a burden exceeding the physiological
circumference of the animal limb and are prone to infection and pin
loosening, especially in long-term studies. Therefore, defect fixation with
plates or internal fixators offers considerable advantages.
124
Advantages Disadvantages
External fixator - Versatility - Ease of application - Marginal effect on
surrounding soft tissue - Minimal intraoperative
trauma - Open space for
implantation of biomedical construct
- Clinically only applied as temporary fixation device
- Schanz screw loosening - Pin track infections
Intramedullary nail - Standard treatment for diaphyseal fractures of lower extremity
- Availability of UTN to avoid reaming related problems
- Central load carrier - High tolerance of
maximum applied forces - Low axial deviations - Reaming debris as
possible source of multipotent stem cells
- Impairment of bone blood circulation by reaming
- Thermal necrosis after reaming
- Air or fat embolism - Failure of locking bolts - Limited application of
load-bearing scaffolds - Prolonged healing period
Plate fixation - Standard treatment for metaphyseal fractures
- Optimal reduction - Minimal influence on
defect (LC-LCP) - Open space for
implantation of biomedical construct
- Eccentrical load carrier - Impairment of periosteal
blood flow (non LCP) - Bone loss through stress
protection (non LCP) - Unclear role of plate
fixation in tibial shaft fractures
- Prone to axial deviations and implant failure
Table 1: Advantages and disadvantages of different fixation devices.
Intramedullary nail versus plate fixation versus external fixation
Diaphyseal tibial fractures are most commonly stabilized by intramedullary
nailing. Despite its wide-spread application, it is associated with both
positive and negative side aspects. Reaming can impair the bone’s blood
circulation system considerably [206, 207]. Reaming and nail insertion
significantly increase the intramedullary pressure. This can lead to air or fat
embolisms causing pulmonary microvascular damage [208, 209] and a rise
in temperature in the medullary canal. Rise in temperature of 50°C and
more have been described [210, 211] which may result in thermal necrosis
of bone tissue also altering the endosteal architecture. This may induce
biological failure compromising fracture or defect healing [212]. The damage
125
of local tissue, however, is mostly reversible and compensated for within 6-8
weeks. With unreamed nailing systems the interlocking screws are prone to
failure while small dimension solid tibial nails cannot always provide
adequate stability. The introduction of angle stable locking bolts could
reduce failure rates. In the context of tissue engineering an intramedullary
nail may impede the placement of a solid, one-piece load-bearing scaffold or
tissue engineered construct. Central load carriers, however, are less
susceptible for tilting in the frontal plane. Therefore, custom-made
intramedullary nailing systems have been applied in a number of
publications for the fixation of segmental bone defects in large animal
models [59, 93, 97, 98] (Fig. 2).
Fig. 2: Schematic representation of commonly applied methods for the fixation of
segmental defects in large animal models.
Metaphyseal fractures are preferably treated with angle-stable plates [213].
Sufficient reduction and stability can be achieved by plate osteosynthesis
even in complicated fractures [214, 215]. Conventional plate osteosynthesis,
126
however, can result in histologically and radiologically verifiable bone loss,
which has been attributed to stress protection (Wolff’s law). Studies have
revealed an increase in cortical bone porosity proximate to the plate soon
after plate osteosynthesis. This porosity was explained by impaired vascular
perfusion underneath the plate [216] due to high compression forces
between plate, periosteum and bone [217]. Contact pressure between plate
and bone and resulting circulatory disturbances may prolong fracture
healing and increase the risk of infection and re-fracture after implant
removal [218]. Therefore, plate systems were developed where load and
torque transmission act through the screws only as achieved by angle-
stable, interlocking screws. Thus, the stabilisation system acts rather like an
external fixator making plate-bone contact unnecessary to achieve stable
fixation. The application of monocortical screws can further minimize screw-
related intramedullary circulatory disturbances [219]. The role of plate
fixation in the treatment of human tibial shaft fractures varies between
different clinics and colleges of surgeons, as the literature lacks randomised
trials comparing plate fixation with other established treatment concepts. In
the area of tissue engineering and related animal studies, defect fixation
with internal fixators offers the great advantage of a minimal influence of the
fixation device on the created defect site both concerning space for scaffold
implantation and biological factors. When compared to external fixators or
intramedullary nails, rates of infections (pin-track infection), infection related
complications and non-union rates (6-25%) are lower. However, higher
numbers of mal-alignment may be observed [220].
Clinically, external fixation is commonly applied as a temporary fixation
127
device in case of severe open or contaminated fractures. In animal models,
external fixators have been widely used offering versatility and ease of
application [3, 12, 96]. Due to minimal intraoperative trauma, the
surrounding soft tissue is only affected marginally with external fixators.
However, Schanz screw loosening is the most common complication with
external fixator often resulting in pin track infections. When compared to
intramedullary nails, external fixators do not affect the defect site as
extensively leaving open space for the implantation of bone grafts and
tissue engineered constructs. However, with external fixators healing
periods are reported to be significantly longer when compared to other
fixation devices [205].
In conclusion, the establishment of a critical size segmental defect model in
a large animal is a challenging task. Hence, in the following section the
knowledge and experience of how to establish such a defect model in an
ovine model (aged 6-8) is described in detail.
128
Roadmap to establish a preclinical model for segmental
bone defect research
To translate scaffold-based bone engineering concepts from bench to
bedside, the level of therapeutic benefit must be conclusively demonstrated
and the essential underlying biological mechanisms elucidated. The
functionality of engineered bone constructs must be validated in a high load-
bearing large animal model and characterized with respect to scaffold
biomechanical competence, integration, degradation & resorption, cell
survival and release kinetics of growth factors (if applicable).
Among the different indications of bone grafting in orthopaedic surgery, in
the current study, focus was placed on critical-sized, diaphyseal, tibial
defects for several reasons:
(i) The tibia is the most commonly fractured long bone, with an
incidence of 17 per 100,000 person-years [221]. Compromised
healing and therefore delayed healing or non-unions in tibial shaft
fractures are common. Up to 60% of segmental defects occur in
the tibia [222]. The overall rate of delayed unions in tibial fractures
ranges from 5% to 61% and the overall rate of non-union might be
as high as 21% [223-226];
(ii) Impaired functionality is significant as the tibia is a highly weight-
bearing bone and tibial shaft fractures often occur in young and
active adults. With prolonged treatments, as in the case of
delayed or non-union, the resultant missed days of work and loss
of wages increase both the direct costs of care and the indirect
costs associated with lost productivity;
129
(iii) Scaffold/cell [12] and scaffold/BMP [115] based concepts have
been used with interesting preliminary results in tibial defects but
there are no randomized studies comparing the results of different
scaffold based treatment concepts to the performance of
autologous bone graft in an ovine segmental defect model.
Pilot study limited contact locking compression plate (LC-LCP)
In order to establish a critical-sized segmental bone defect model in the
sheep tibia, a pilot series of 3 animals (merino sheep, weight 39±1 kg, age
6-7 years) was conducted. All animals were in good health but of small
statue (related to the geographic and climatic conditions in central
Queensland, Australia). Under general anaesthesia, a 20 mm mid-
diaphyseal defect was created using an oscillating saw. A previous literature
search had shown that a 20 mm defect was the smallest critical-sized defect
described in sheep tibiae [126]. The defect was stabilized using a 9-hole 4.5
mm narrow LC-LCP (limited contact locking compression plate, Synthes)
anteromedially with three bicortical screws proximal and distal and left
untreated. The implant, a modern internal fixation system, was chosen since
it had previously been used with success to stabilize an experimental
fracture in sheep femora (Wullschleger et al., unpublished data). After
surgery, the operated leg was bandaged with hard plaster (Vet lite, Runlite
SA, Micheroux, Belgium). The animals were held in a suspension trolley for
24 h to allow recovery from anaesthesia prior to release into a paddock. The
animals were allowed unrestricted weight-bearing. In all three animals, 7-10
days after surgery, plate bending occurred as suspected by clinical
130
assessment and confirmed by x-ray analysis (Fig. 3). It was concluded that
these failures resulted from critical loads (valgus stress) occurring during the
act of the animal standing up from a lying position. It should be noted
however that while the implant was proven not to be strong enough in an
immediately fully weight bearing animal model, this should not be
interpreted as having implications for the treatment of similar conditions in
human patients, since compliant patients can be advised to weight bear only
partially and to avoid critical loads that may cause implant failure.
Fig. 3: 2 cm segmental bone defect in a sheep tibia stabilized with a 9-hole narrow
4.5 mm limited contact locking compression plate (LC-LCP plate) (A, E, F); implant
bending at day 7 due to critical loads and valgus stress (B). Starting defect
regeneration was observed at day 14 (C) and the uncontrolled biomechanical
stimulation led to defect bridging on the postero-lateral side at day 28 (D).
131
Finite element modelling
In response to the in vivo failures of the internal fixation plate seen in the
first pilot study, finite element analysis was conducted to aid the selection of
a fixation device with mechanical properties sufficient to resist failure of the
plate via bending. In addition to the material properties and the bulk size of
the plates, the type of screw hole, single or combination, and screw hole
configuration are determinants of implant stiffness. Four plates with different
sizes and screw hole type were investigated, 3.5 mm broad LCP (stainless
steel, 220 GPa), 4.5 narrow LCP (stainless steel, 220 GPa), 4.5 mm broad
LCP (titanium, 105 GPa), and the LISS plate (less invasive stabilization
system, titanium alloy, 110 GPa).
Finite element models (tetrahedral elements) were constructed for each of
the four plates from technical drawings supplied by the manufacturer
(ABAQUS v6.5). Four-point bending tests were then simulated to determine
the equivalent bending stiffness (ISO 9585: Implants for surgery –
Determination of bending strength and stiffness of bone plates) of each of
the plates alone using a linear elastic solver. The 4.5 narrow LCP used in
the first pilot study was found to have the lowest equivalent bending
stiffness of all the plates tested. The 4.5 mm broad LCP was determined to
be the stiffest plate (15.32 Nm2) closely followed by the LISS plate (14.81
Nm2).
132
Implant testing
The most rigid plates determined by finite element modelling were selected
for in vitro biomechanical testing. In addition to the locking plates, a dynamic
compression plate was also included. In vitro biomechanical testing was
performed to determine the stiffness of the implant-bone constructs. A four-
point bending test was carried out to determine the equivalent bending
stiffness. Briefly, a defect of 2 cm was created in ovine cadaver tibiae and
fixed with either (n=3) a 4.5 mm narrow LC-LCP (stainless steel), a LISS
(titanium) and a 4.5 mm broad DCP (dynamic compression plate, stainless
steel)(all Synthes, Fig. 4). N.B. The broad screw hole arrangement of the
4.5 mm broad LCP was found to be unsuitable for the narrow sheep tibia
and was thus substituted with the LISS plate, which showed similar bending
stiffness to the 4.5 mm broad LCP in finite element analyses. Both ends of
the tibiae were potted in cups for load transfer using bone cement (poly-
methyl-methacrylate (PMMA)) in a standardized manner. To ensure that the
same moment arm acted on the respective constructs care was taken to
ensure that the free length between the cups remained constant for all
bones. Constructs were mounted in four-point bending configuration and
loads applied using a biaxial universal testing machine (Instron 8874,
Instron, Norwood, USA). The test was repeated three times for each
bone/implant construct. During the entire procedure, the tibiae were
wrapped in saline soaked gauze to avoid the samples drying out. The
equivalent bending stiffness was calculated according to ISO norm 9585
“Implants for surgery – Determination of bending strength and stiffness of
bone plates”, which specifies the test configuration based on the plate screw
133
hole spacing. The highest stiffness was determined for the 4.5 mm broad
DCP with E=0.058 Nm2 (SD=0.0075). The modified LISS plate provided only
marginally higher stiffness (E = 0.028, SD = 0.0049) than the 4.5 narrow LC-
LCP used in the pilot study (E=0.023 Nm2, SD=0.0024)(Fig. 5).
Fig. 4: The figure illustrates the implants used to biomechanically determine the
stiffness of implant-ovine bone constructs in vitro: A 4.5 mm narrow Limited
Contact Locking Compression Plate (LC-LCP, A, D), a tibial Less Invasive
Stabilization System (LISS, B, E) plate, and a 4.5 mm broad Dynamic Compression
Plate (DCP, C, F)(all Synthes). The length of the LISS plate was tailored to fit an
ovine tibia as indicated by the red bar. G and H demonstrate the setup of the four
point bending test.
134
Fig. 5: The figure illustrates the equivalent bending stiffnesses as per ISO 9585:
Implants for surgery – determination of bending strength and stiffness of bone
plates. The highest stiffness was determined for the 4.5 mm broad DCP with
E=0.058 Nm2 (SD=0.0075). Error bars represent standard deviations.
Pilot study dynamic compression plate
For further experimental surgeries, the 4.5 mm broad DCP as the stiffest
implant was chosen. Another series of three animals (42±1 kg, age 6-7
years) was then operated on to evaluate if the new implant would show
sufficient biomechanical strength in vivo and to evaluate if a defect size of 2
cm was critical. The 10-hole DCP plate was fixed with 4 screws proximal
and 3 screws distal of the defect. Care was taken to completely remove the
periosteum within the defect area as well as 1.5 cm proximal and distal of
the defect [185]. Animals were subjected to the same postoperative
treatment as described previously. After 12 weeks, two out of three animals
showed islets of bridging of the defect as assessed by conventional x-ray,
!CT analysis (microCT 40, Scanco, Switzerland) and histology. For !CT
analysis, samples were scanned with a voxel size of 16 !m. Samples were
135
evaluated at a threshold of 220, a filter width of 0.8 and filter support of 1.0.
No further implant failures in terms of plate bending had occurred. In one of
the animals where the defect was bridged, the most proximal screw had
loosened (Fig. 6). This may have changed the rigidity of fixation and hence
the biomechanical conditions allowing micro movements at the defect site
which in turn could have stimulated callus formation. Other factors such as
surgical technique and inter animal variability in terms of weight etc. might
have caused the second of three defects to bridge.
Fig. 6: 2 cm ovine tibial defect stabilized with a broad 4.5 mm dynamic
compression plate (DCP) (A) fixed with 4 bicortical screws proximally and 3 screws
distally. X-ray image after surgery (B) and after 12 weeks (C) shows loosening of
two screws proximal of the defect (arrows) and consecutive bone formation.
To ensure a non-union rate of 100%, the defect size was then increased to
3 cm, again the periosteum was removed. This resulted in non-unions after
12 weeks in all of eight animals in the negative control group (weight 45±2
136
kg, age 6-7 years). Hence, a defect size of 3 cm in the sheep may be
considered of critical size (Fig. 7). As a positive control eight 3 cm defects
were reconstructed with autologous cancellous bone graft from the iliac
crest representing the gold standard treatment. No implant failures were
observed and after 12 weeks all defects had healed and radiographic
evaluation revealed defect consolidation. However, it must be pointed out
that radiographs cannot provide any information on bone quality and
functionality, thus it was necessary to undertake biomechanical testing. This
baseline data was later on used to assess the performance of medical grade
polycaprolactone-tricalcium phosphate (mPCL-TCP) scaffolds in
combination with and without human recombinant bone morphogenetic 7
(rhBMP-7) in respect to their bone regeneration capacity.
Fig. 7: 3 cm mid-diaphyseal tibial defect in a sheep tibia (A) stabilized with a 4.5
mm dynamic compression plate (DCP) (B). Radiographic images immediately (C)
and 12 weeks after surgery (D) show the critical nature of the defect size.
137
Summary
To replace or restore the function of traumatized, damaged or lost bone is a
major clinical and socio-economic challenge. Bone tissue engineering has
been suggested as an alternative strategy to regenerate bone. In principal,
the discipline of tissue engineering aims at combining cells with scaffolds, or
appropriate growth factors such as FGF-2 (fibroblast growth factor 2), VEGF
(vascular endothelial growth factor) or BMPs (bone morphogenetic
proteins), to initiate and stimulate tissue repair and regeneration. Despite
initial success, it must be recognized that bone engineering has not yet
been able to deliver significant progress in terms of translated clinical
applications and commercialized products. To tackle major bone tissue
engineering problems, researchers must perform functional assessment of
the biological and biomechanical parameters of generated constructs.
Furthermore, to allow comparison between different studies, animal models,
fixation devices, surgical procedures and methods of taking measurements
need to be standardized to achieve an efficient accumulation of reliable data
as a foundation for future orthopaedic and tissue engineering developments
and their translation to the clinic.
Conventional DCPs may not meet all the requirements postulated for a
modern fracture fixation device and higher rates of screw loosening have
been described (full load carried by the screws) when compared with LCPs
where loads are equally carried by both screws and plate which, from a
mechanical point of view, can be seen as one unit as a result of the locking
mechanism. It was reported that DCPs impede periosteal circulation through
compression other studies, however, have vitiated these findings.
138
Nevertheless, DCPs are still used in clinical settings. To minimize effects on
periosteal blood flow, limited contact (LC) plates were developed and
introduced to the orthopaedic/trauma market [227] [228]. To further preserve
biomechanical integrity, the concept of locked internal fixators was
introduced offering improved long-term biomechanical stability and
decreased susceptibility to infection. However, narrow LCPs do not provide
the required stiffness in an immediate fully load-bearing animal model and
will only be of sufficient strength if the animals are immobilized for a longer
period of time (6-12 weeks) until the regenerating bone has gained some
intrinsic biomechanical stability [229]. Broad LCPs on the other hand do not
fit ovine tibiae as a result of their dimensions and therefore do not represent
an alternative.
In conclusion, the most important issues related to the establishment of a
large preclinical model for segmental bone defect research were addressed
and discussed, and it was demonstrated how to develop such a model
relying on a DCP plate fixation system. With respect to the “Valley of Death”,
an important milestone was achieved in establishing a highly reproducible
large animal model, which is an essential requirement to systematically
assess different bone grafting materials, scaffolds, tissue engineered
constructs and growth factors on the path towards the translation into a
routine clinical application.
139
Chapter V
Reconstructing large segmental bone defects in an ovine model by
tissue engineering methods
140
141
Introduction
In orthopaedic and trauma surgery, extensive bone loss is associated with
major technical and biological problems. Fractures in cancellous bone areas
- such as the proximal humerus, the distal radius or the tibial plateau - often
lead to impaction of bone and subsequent defect formation after reduction.
Other causes of traumatic bone loss include highly comminuted fractures
and atrophic diaphyseal non unions of long bones. Bone grafts as the gold
standard treatment possess osteoconductive and osteoinductive properties,
however, they still face significant disadvantages. These include limited
access and availability, donor site morbidity and haemorrhage, increased risk
of infection, and insufficient transplant integration with following graft
devitalisation and subsequent resorption resulting in decreased mechanical
stability. As a result, recent research focuses on the development of
alternative therapeutic concepts.
Engineering bone by combining a suitable scaffold with osteoblastic or
osteoprogenitor cells and/or bone formation stimulating growth factors has
advanced as a promising new approach to bone repair and regeneration.
However, in order to establish a tissue-engineering concept in a clinical
setting, a rigorous demonstration of the level of therapeutic benefit in
clinically relevant animal models is absolutely essential. Difficulties in
integrating individual technical discoveries in model tissue engineering
systems and in manufacturing scale up combined with shortages in funding,
and in regulatory approval that are associated with the translation of tissue
142
engineering based concepts make clinical applications a difficult target to
reach [230].
To circumvent the disadvantages associated with bone grafts, a variety of
different materials for bone replacement have been developed [231].
Calcium phosphates (CaP) form a large group of materials that are
considered suitable substitute biomaterials for bone augmentation. Calcium
phosphate based materials have been shown to be biocompatible evoking
no adverse inflammatory reactions after implantation [6, 232-234]. Calcium
phosphate has also demonstrated osteoinductive effects in numerous
publications [6, 234].
Aliphatic polyesters such as PLLA/PDLA, PLLLA/PGA, PCL and others have
demonstrated excellent safety profiles in multiple in vitro, animal, and clinical
studies and have approval for clinical applications [235, 236]. They do not,
however, meet the structural and mechanical requirements for use in
orthopaedic applications of the lower extremities. Therefore, a second
generation of scaffolds has been developed. These scaffolds are based on a
medical grade PCL-CaP composite [160, 237], and combine the favourable
mechanical properties of the polyester with the osteoinductive characteristics
of the ceramic component. So far, these mPCL-TCP scaffolds have been
tested in rat and pig skull defect as well as in a spinal fusion model with
promising outcomes [238].
Healing of osseous tissue is regulated by growth factors and other cytokines
in a sequence of overlapping events similar to cutaneous wound repair. In
ideal circumstances, this process mimics embryonic bone development,
143
allowing the replacement of damaged bone with new bone, rather than with
fibrous scar tissue. This process is driven by cellular and molecular
mechanisms controlled by the TGF-" superfamily of genes, which encodes a
large number of extracellular signalling growth factors [239]. Bone
morphogenetic proteins (BMPs) are a well-studied group of these growth
factors involved in the processes of bone healing; and the human genome
encodes at least 20 of these multifunctional polypeptides [240]. Among
several of its functions, BMPs induce the formation of both bone and
cartilage by stimulating the cellular events of mesenchymal progenitor cells.
However, only a subset of BMPs, most notably BMP-2, -4, -7, and -9, have
osteoinductive activity, a property of inducing de novo bone formation [241].
Studies involving mutations of BMP ligands, receptors and signalling proteins
have shown important roles of BMPs in embryonic and postnatal
development. Severe skeletal deformation, development of osteoporosis,
reduction in bone mineral density and bone volume are all aberrations
associated with disrupted and dysregulated BMP signalling [242, 243].
Several other growth factors produced by osteogenic cells, platelets and
inflammatory cells participate in bone healing, including IGF-1 and -2, TGF-
"1, PDGF and FGF-2 [244]. The bone matrix serves as a reservoir for these
growth factors and BMPs, which are activated during matrix resorption by
matrix metalloproteases [245]. Additionally, the acidic environment that
develops during the inflammatory process leads to activation of latent growth
factors [246], which assist in the chemo-attraction, migration, proliferation,
and differentiation of mesenchymal cells into osteoblasts or chondroblasts
[246]. All of these functions are driven by a complex mechanism of
144
interaction among growth factors and other cytokines, which are influenced
by several regulatory factors.
The majority of studies investigating the effects of BMP-7 on new bone
formation were conducted in small animal models and experimental studies
in critical-size defect models of the weight-bearing lower extremity show non-
uniform results. Therefore, in this present study, an animal study was
designed to investigate the effects of rhBMP-7 in combination with mPCL-
TCP composite scaffolds built by computer aided design using rapid
prototyping technologies on bone healing. It was hypothesized that such a
tissue engineered bone graft may enhance the treatment of large segmental
bone defects without the need of vascularised autografts, non-vascularised
autografts, and/or allografts, and could therefore represent a clinical
alternative to bone autografts for the reconstruction of large tibial and femoral
defects.
145
Material and Methods
Scaffold fabrication and preparation
Bioresorbable cylindrical scaffolds of medical grade )-poly-caprolactone
incorporating 20% "-tricalcium phosphate (mPCL–TCP) (outer diameter: 20
mm, height: 30 mm, inner diameter: 8 mm) (Fig. 1) were produced by fused
deposition modelling (FDM) as described previously (Osteopore
International, Singapore; www.osteopore.com.sg) [169]. The structural
parameters of the scaffolds were tailored by computer aided design (CAD)
and included 100% pore interconnectivity within a range of 350–500 µm size,
70% scaffold porosity, and a 0/90° lay down pattern [169] [247]. This
architectural layout is particularly suitable for load bearing tissue engineering
applications since the fully interconnected network of scaffold fibres can
withstand early physiological and mechanical stress in a manner similar to
cancellous bone [169]. Moreover, the architectural pattern allows retaining of
coagulating blood during the early phase of healing, and bone in-growth at
later stages. Prior to surgery, all scaffolds were surface treated for six hours
with 1M NaOH and washed five times with PBS to render the scaffold more
hydrophillic. Scaffold sterilization was achieved by incubation in 70% ethanol
for 5 min and UV irradiation for 30 min.
146
Fig. 1: 3D !CT reconstructions of a cylindrical mPCL-TCP scaffold produced via
fused deposition modelling with an outer diameter of 2 cm and a height of 3 cm.
Scaffold parameters include a porosity of 70%, fully interconnected pores of a
diameter of 350-500 !m, and a 0/90˚ lay down pattern. Top view (A) and lateral view
(B).
To load the scaffolds with OP-1 implant 3.5 ml of sterile saline were added to
one vial containing 3.5 mg recombinant human BMP-7 and 1 g of bovine
type I collagen and thoroughly mixed. Then, the resulting putty was
transferred to the inner duct of the scaffold and the contact interfaces
between bone and scaffold (Fig. 2).
Fig. 2: To load the scaffolds with OP-1, the lyophilized protein was mixed with 3.5 ml
of sterile saline (A, B) and transferred to the inner duct of the scaffold and onto the
contact interfaces between bone and scaffold (C, D).
147
Animal surgery
Anaesthesia and pre-operative treatment
A jugular venous line was installed under aseptic conditions, and 15-20 ml of
1% propofol were injected to induce general anesthesia. The respective
animal was then intubated with a 9-10 mm silicon endotracheal tube and
connected to an automatic respirator (Campbell anaesthetic ventilator) for
assisted ventilation with 2l O2/min. The general anesthesia was maintained
with propofol at a rate of 120-140 ml/min. For analgesia, Buprenorphine (0.1
mg per 10 kg body weight) was administered, for antibiotic prophylaxis
gentamycine (5 mg/kg) and cephalothin (25 mg/kg). The animal's ECG, heart
rate, oxygen saturation and end-tidal carbon dioxide levels were monitored
and recorded continuously.
Defect model
Animals (Merino sheep, average weight: 42.5 kg, age: 6-7 years) were
placed in right lateral recumbency. The right hindlimb was carefully shaved
and thoroughly desinfected with 0.5% chlorhexidine solution red in 70 %
ethanol. The animal torso and surroundings were then covered with sterile
sheets, the surgical area additionally with Opsite (Smith and Nephew). The
right tibia was exposed by a longitudinal incision of approximately 12 cm
length on the medial aspect of the limb. A bone fixation plate (4.5 mm broad
DCP, 10 holes, Synthes) was adjusted to the morphology of the bone by
bending (plate-bending press, Synthes) and applied to the medial tibia. The
distal end of all plates was placed exactly 2.5 cm proximal of the medial
148
malleolus. The screw holes were drilled and the plate was fixed temporarily
with 2 screws adjcacent to the anticipated defect (Fig. 3 A).
The middle of the defect site was measured and marked with a raspatory;
afterwards, the plate was removed. A distance of 1.5 cm each proxiamally
and distally of the defect middle was measured and marked to define the
osteotomy lines (Fig. 3 B, C). Next, the soft tissue inserting to the bone in the
designated defect area was detached and a wet compress was placed
between bone and posterolateral soft tissue to avoid damage to proximate
nerve and blood vessels during osteotomy (Fig. 3 D). Parallel oteotomies
perpendicular to the bone’s longitudinal axis were performed with an
oscilating saw (Stryker) under constant irrigation with saline solution to
prevent heat induced osteonecrosis whilst the bone segment of 3 cm length
was excised (Fig. 3 E). Care was taken to completely remove the periosteum
within the defect area and 1.5 cm proximally and distally of the osteotomy
lines (Fig. 3 F). The bone fragments were realligned and fixed applying the
dynamic compression plate (DCP) with 4 screws proximally and 3 screws
distally to leave a defect gap of exactly 3 cm size (Fig. 3 E). The wound was
closed in layers with a 2-0 Monocryl (Ethicon) and a 3-0 Novafil (Syneture)
suture for the skin. The closed wound was sprayed with Opsite (Smith and
Nephew), covered with pads and bandaged (Vetrap, 3M). After recovery
from anaesthesia, animals were allowed unrestricted weight bearing.
149
Fig. 3: To create a 3 cm segmental defect, the plate was temporarily fixed with two
screws (A), the screw holes were drilled, the defect middle and osteotomy lines
were marked (B, C), and the bone segment was removed after osteotomy (D, E).
Care was taken to remove the periosteum (F) before the bone fragments were
realigned and fixed with plate and screws (G).
150
Harvest of autologous cancellous bone graft
Autologous, cancellous bone graft was harvested from the left iliac crest. The
surgical area was shaved and desinfected with 0.5% chlorhexidine red in
70% ethanol. A 5 cm incision was made following the iliac crest (Fig. 4 A),
the inserting musculature was carefully detached and the cortical bone of the
lateral os ileum was fenestrated (2 x 2 cm) using a hammer and osteotome
(Fig. 4 B). Care was taken not to fracture the ala ossis ilii. The resulting lid
was carefully removed with a raspatory (Fig. 4 C) and the cancellous bone
harvested utilizing a bone curette (Fig. 4 D). The lid was then reinserted (Fig.
4 E), and the musculature reattached with 2-0 Vicryl sutures (Ethicon), and
the wound closed in layers (Fig. 4 F). The closed wound was sprayed with
Opsite (Smith and Nephew).
Fig. 4: The figure illustrates the harvesting procedure for autologous bone grafts
with the incision (A), the fenestration of the os ilium (B), removal of the lid (C),
harvested graft (D), lid reinsertion (E) and wound closure (F).
151
Experimental groups
Experimental groups of the ovine segmental bone defect study included
empty control defects as a negative control, defects reconstructed with
autologous cancellous bone graft from the iliac crest as a positive control,
defects reconstructed with mPCL-TCP scaffolds and defects augmented with
mPCL-TCP scaffolds and 3.5 mg of OP-1 (OP-1 Implant, Stryker)(Table 1).
OP-1 or recombinant human bone morphogenetic protein 7 (rhBMP-7) was
introduced as an additional, biologically active, osteoinductive element.
Table 1: Experimental groups included in the ovine segmental bone defect study.
Euthanasia
The animals were euthanized by intravenous injection of 60 mg/kg
pentobarbital sodium (Lethabarb, Virbac, Australia). After euthanasia, both
hind limbs were exarticulated at the knee and the tibia was dissected and the
implanted osteosynthetic material was removed. The samples were then
processed and analysed.
152
Radiographic analysis
Immediately after surgery, after 6 and 12 weeks, conventional x-ray analysis
(3.2 mAs; 65 kV) in two standard planes (anterior-posterior and medial-
lateral) was performed to assess bone formation.
Computed tomography
After sacrifice, a clinical CT scanner (Philips Brilliant CT 64 channels) was
used to scan the experimental limbs. A dipotassium phosphate phantom was
used to calibrate measurements of mineral density.
3D reconstructions from the CT data were generated with AMIRA® 5.2.2
(Visage Imaging GmbH, Berlin, Germany) (Visage Imaging) with a threshold
of 300 and qualitative analysis was performed to assess mineralization within
the defect and bridging. A scoring system (Fig. 5) was developed to describe
callus formation within the defect zone (1 point for bone formation in the
proximal defect half, 1 point for bone formation in the distal defect half),
external callus formation (1 point for external callus in the proximal, 1 for
external callus in the distal defect half; additional 2 points for external callus
in more than one location), and defect bridging (2 points).
153
Fig. 5: 3D reconstruction of a 3 cm defect after 12 weeks overlayed with the scoring
system
For quantitative analysis, the CT datasets of operated and contralateral intact
tibia of each animal were first cropped to image stacks with equal bounding
box dimensions using AMIRA®. Next, cortical bone and callus tissue were
segmented by choosing appropriate threshold values (lower threshold: 300)
for the measured grey levels. A 3D surface was generated and saved as a
binary file (STL binary Little Endian format). These .stl files were loaded into
Rapidform2006 (Inus Technology, Seoul, Korea) and a minimum of 4
corresponding reference points was selected on each intact and defect tibia
and bound to the respective shell. Intact and defect tibia were then registered
to align their shells utilizing the previously defined common geometries
between them. The reference point coordinates of the defect tibia were
recorded prior to and after registration. The coordinates of the initial and final
points were entered in an in-house MATLAB program (MATLAB 7.6.0,
154
MathWorks, Inc., Natick, MA, USA) to determine the matrix for the required
transformation. This transformation matrix was then used to align the image
data stacks in AMIRA®. This alignment resulted in intact and defect tibia to
have the same orientation. Next, the amount of newly formed bone in three
defined regions/volumes of interest within the 3 cm defect area was
calculated utilizing an in-house MATLAB program (Fig. 6).
Fig. 6: CT DICOM image of an intact ovine tibia (axial view) defining the three
regions of interest.
Biomechanical testing
Both ends of the tibiae were embedded in 80 ml Paladur (Heraeus-Kulzer
GmbH) and mounted in an Instron 8874 biaxial testing machine (Fig. 7). By
leaving as much soft tissue as possible attached, bone samples were
prevented from drying out.
A torsion test was conducted at an angular velocity of 0.5 deg/s and a
compressive load of 0.05 kN until the fracture point was reached (right tibiae
counter clockwise, left tibiae clockwise). The contralateral tibia was used as
a paired reference. The torsional moment (TM) and torsional stiffness (TS)
155
values were calculated from the slope of the torque-angular displacement
curves and normalized against the values of the intact contralateral tibiae.
Fig. 7: Potting of a sample tibia for biomechanical testing. First, the proximal part
was embedded in Paladur (A) with the tibial axis vertically aligned. The tibia was
then rotated (B) and the distal part was embedded (C).
!CT analysis
For !CT analysis, a region of interest including the 30 mm defect gap and 5
mm of adjacent bone each proximally and distally was selected. Samples
were scanned (vivaCT 40, Scanco medical) with a voxel size of 19.5 !m.
Samples were evaluated at a threshold of 220, a filter width of 0.8 and filter
support of 1.0 and analysed for bone volume, bone mineral density, and
trabecular thickness within the defect using the software supplied by the
manufacturer of the !CT. In addition, the polar moment of inertia (pMOI) in
m4 was calculated according to the following formula: Jz = ! "! where Jz
represents the polar moment of intertia about an axis z, dA and elemental
area, and # the radial distance to the element dA from the axis z [248] (Fig.
156
8). The pMOI represents a quantity used to predict an object’s ability to resist
torsion.
Fig. 8: A schematic showing how the polar moment of inertia is calculated for an
arbitrary shape about an axis o. # is the radial distance to the element dA.
157
Histology
After biomechanical testing, specimens were trimmed to 15 cm length and
fixed in 10% neutral buffered formalin. For histological analysis, the mid-
defect regions were sectioned in the transversal and sagittal plane (Fig. 9).
Callus tissue composition was evaluated on 6 µm-thick methylmethacrylate
(Technovit 9100 NEU, Heraeus Kulzer, Germany) embedded sections
(embedding procedure according to the manufacturer’s protocol), stained
with Safranin Orange/von Kossa (mineralized tissue, black), and Movat’s
pentachrome [249] to demonstrate bone (yellow), cartilage (deep green) and
fibrous tissue (light green-blue).
Fig. 9: The schematic illustrates the frontal and sagittal cutting planes used for
histological sectioning and their spatial arrangement in respect to the defect area.
Statistical analysis
Statistical analysis was carried out using a two-tailed Mann-Whitney-U-test
(SPSS 16.0, SPSS Inc.) and p-values were adjusted according to Bonferroni-
Holm. Sample size for the study was determined with a power analysis
performed on the torsional strength data reported for a comparable critical
size defect study in sheep tibiae [93].
158
Results
Animal model
A surgical technique was successfully developed to excise a 3 cm mid-
diaphyseal tibial bone segment without affecting surrounding soft tissue,
blood vessels and nerves. In addition, a single sided ABG harvesting
procedure was followed that ensured sufficient amounts of graft for defect
reconstruction. No postoperative infections or other complications were
observed. The chosen 4.5 mm broad DCP was proven to be biomechanically
sufficient to prevent implant failure. All animals were in good health and
survived the experimental period gaining weight in the months following
surgery. By ensuring complete removal of the periosteum within the defect
area, a defect model of critical nature could be established showing no
defect bridging in all animals included in the empty defect control groups.
X-ray analysis
Conventional X-ray analysis in two planes after 12 weeks confirmed the
critical nature of the defect. None of the empty control defects (n=8) showed
signs of external callus formation and only marginal bone formation within
the defect area. Only minor external callus and bone formation within the
defect was observed in the scaffold group. Callus was located mainly
postero-laterally. However, full defect bridging had occurred in all defects
reconstructed with ABG or mPCL-TCP with rhBMP-7. ABG treatment did not
result in substantial external callus formation, but development of new bone
within the defect. RhBMP-7 in combination with a mPCL-TCP scaffold had
stimulated both external callus formation (mainly postero-laterally) and bone
159
formation within the defect gap (Fig. 10). In all groups, bone formation was
observed along the stainless steel plate.
Fig. 10: Representative x-ray images after 12 weeks of an empty control defect (A),
a defect reconstructed with cancellous bone graft from the iliac crest (B), a defect
augmented with mPCL-TCP + rhBMP-7 (C), and a defect treated with a mPCL-
TCP scaffold (D). The images show clear radiographic signs of defect bridging for
the autograft and rhBMP-7 group, with external callus formation in the rhBMP-7
group. No bone formation was observed within the empty control defect and only
little bone formation for the scaffold only group.
160
Computed tomography
Qualitative CT analysis after 12 weeks confirmed the critical nature of the
defect showing non unions in all eight empty control defects, which were
filled with soft tissue only. Only minor bone formation was observed in the
mPCL-TPC scaffold group, mainly on the posterolateral side. However, full
defect bridging had occurred in all defects reconstructed with ABG or mPCL-
TCP + rhBMP-7 (Fig. 11). No radiographic signs of inflammation were found.
Scaffolds showed good osseointegration without any signs of resorption. The
qualitative CT score performed on 3D reconstructions (AMIRA 5.2.2,
threshold of 300) (Fig. 12) of all samples assessing bone and callus
formation showed significant differences between ABG (mean=6.625 out of 8
possible points) or mPCL-TCP + rhBMP-7 (mean=7.125) when compared
with the empty control (mean=3.25) or scaffold only group. Moreover, a
significant difference in bone formation between empty defect and scaffold
only group was found. No statistically significant difference was found
between ABG and mPCL-TCP + rhBMP-7 group (Fig. 13, 14).
161
Fig. 11: Union rates (A) and qualitative CT score (B, box plot) assessing bone
formation within the defect, external callus formation and defect bridging (n=8).
Asterisks indicate statistical significance.
A
B
162
Fig. 12: Representative 3D CT data reconstructions (AMIRA 5.2.2) of critical
segmental bone defects, which were left untreated (A), reconstructed with ABG (B),
an mPCL-TCP scaffold (C) or an mPCL-TCP scaffold combined with rhBMP-7 (D).
Median values of total bone volume (BV) in the defect area were significantly
higher in the rhBMP-7 group (8.6 cm3) than in all other groups (Fig. 10 A). In
the cortical region (region 1) no significant difference between rhBMP-7 (2.96
cm3) and ABG (2.09 cm3) group was found (Fig. 10 B). Bone formation in
region 2, the marrow region, was calculated to be significantly higher in the
ABG group (1.66 cm3) when compare to all other groups (Fig. 11 A). In
contrast to the total bone volume, bone formation in the periosteal region
(region 3) was significantly higher in both the scaffold (1.65 cm3) and rhBMP-
7 group (4.55 cm3) than in the ABG group (0.77 cm3)(Fig. 11 B).
A B C D
163
Fig. 13: Box plot demonstrating the median ± 1st and 3rd quartile (n=8). The figure
illustrates the total bone volume (BV)(A) and bone volume formed in the cortical
region (region 1)(B) after 12 weeks. Asterisks indicate statistical significance. Error
bars represent minimum and maximum values.
A
B
164
Fig. 14: Box plot (the median ± 1st and 3rd quartile) demonstrating the bone volume
formed within the marrow region (A) and periosteal region of the bone (B) after 12
weeks (n=8). Error bars represent minimum and maximum values. Asterisks
indicate statistical significance. Error bars represent minimum and maximum values.
A
B
165
Biomechanical testing
None of the empty control defects showed signs of bridging and were filled
with soft tissue only. Therefore no biomechanical testing could be carried out
on these specimens. Biomechanical testing (Fig. 15 and 16) revealed a
significant higher torsional moment (TM) for the ABG (median=11.19 %) and
rhBMP-7 (14.41 %) groups when compared to the mPCL-TCP group (4.96
%). No significant difference was found between the ABG and rhBMP-7
group. Similar results were observed for torsional stiffness (TS) values with
significantly higher TS values for the ABG (19.317 %) and rhBMP-7 group
(25.043 %) when compared to the scaffold only group (2.535 %). Again no
statistically significant difference was found between the ABG and rhBMP-7
treated group. However, the rhBMP-7 treatment tended to result in higher
values for both torsional moment and stiffness.
166
Fig. 15: Box plot demonstrating median values of torsional moment (A) ± 1st and 3rd
quartile and these values relative to the contralateral tibia (B)(n=8). Error bars
represent maximum and minimum values. Asterisks indicate statistical significance.
A
B
167
Fig. 16: Box plot demonstrating median values of torsional stiffness (A) ± 1st and 3rd
quartile and these values relative to the contralateral tibia (B)(n=8). Error bars
represent maximum and minimum values. Asterisks indicate statistical significance.
A
B
168
µCT analysis
MicroCT analysis confirmed results from the clinical CT scans regarding
union rates and the amount of new bone formation. All of eight defects in the
autograft and scaffold-rhBMP-7 group showed solid bony union after twelve
weeks (Fig. 17). Highest median values of newly formed bone were found for
the scaffold-rhBMP-7 group (median=2254.04 mm3). They did, however, not
differ significantly from values obtained for the positive control, the autografts
(median=2158.77 mm3), as a considerable variation in bone neoformation
was observed (minimum=1126.56 mm3, maximum=4863.87 mm3)(Fig. 18).
Fig. 17: MicroCT sections (A-D) of central defect portions and 3D reconstructions of
3 cm tibial defects (E-H) 12 weeks after surgery. Defects were left untreated (A, E),
reconstructed with autologous cancellous bone from the iliac crest (B, F), a mPCL-
TCP scaffold (C, G) or scaffold plus 3.5 mg rhBMP-7 (D, H). Defect bridging was
found in all specimens of each the autograft and rhBMP-7 group. Scaffolds showed
good osseointegration at the host bone interface.
169
When compared to autograft or rhBMP-7 treated lesions, significantly lower
bone formation was observed in the defects treated with scaffolds only
(median=681.90 mm3) or untreated control defects (median=171.91)
mm3)(Fig. 18). In any case, newly formed bone was still less compared to the
amounts determined for the same anatomic level of the contralateral hind
limbs. The amount of new bone was evenly distributed throughout the
proximal, middle and distal third of the defect although a tendency towards
higher amounts in the proximal defect third was observed (Fig. 19).
The mineral density of the newly formed, woven bone or tissue was found to
be homogenous in the different experimental groups (Fig. 20). Notably,
bone/tissue mineral density values were significantly lower than those
determined for the compact bone of the contralateral healthy tibiae with
median values ranging from 71.37 to 76.34 %.
Analysis of trabecular thickness within the newly formed bone and the
distribution along the z axis showed significantly thicker bone trabeculae in
the proximal third of the ABG group (median=339.65 !m) when compared to
the defect middle (median=257.3 !m) and distal regions (median=246.7 !m)
or corresponding proximal defect segments of the groups involving mPCL-
TCP scaffolds (Fig. 21).
Values for the polar moment of inertia, which describes an object’s ability to
resist torsion, reflected the tendencies of the biomechanical testing results
and determined bone volumes (Fig. 22) and correlated well with these
findings (correlation coefficients: 0.911 and 0.896).
170
Fig. 18: Box plot demonstrating median amounts of newly formed bone in mm3 ± 1st
and 3rd quartile within the 3 cm defects 12 weeks after surgery. Significantly more
bone formation was seen in defects reconstructed with either autograft (ABG) or
scaffold combined with rhBMP-7 (mPCL-TCP+rhBMP-7) when compared to scaffold
alone (mPCL-TCP) or untreated defects. Error bars represent maximum and
minimum values. Bars with asterisks indicate statistical significance.
Fig. 19: Box plot demonstrating the distribution of newly formed bone in mm3 along
the z axis of regenerated defects. A tendency towards higher amounts of bone in
the proximal defect third was observed throughout the different experimental
groups. Error bars represent maximum and minimum values.
171
Fig. 20: Box plot demonstrating median values for tissue mineral density within the
defects ± 1st and 3rd quartile 12 weeks after surgery. Tissue density did not differ
significantly between the experimental groups. Tissue density in the defect zones
was however only about 70% of that determined for the compact bone of the
contralateral limb. Error bars represent maximum and minimum values.
Fig. 21: Box plot illustrating the distribution of trabecular thickness along the z axis
of regenerated defects. Significantly thicker bone trabeculae were observed in the
proximal third of the autograft (ABG) group when compared to the defect middle and
distal regions or proximal defect segments of the groups involving scaffolds.
172
Fig. 22: Box plot demonstrating median values for the polar moment of inertia within
the defect region ± 1st and 3rd quartile 12 weeks after surgery (A) and these values
relative to the contralateral tibia (B). Error bars represent maximum and minimum
values.
A
B
173
Histology
Von Kossa staining on 6 µm-thick methylmethacrylate embedded sections
showed extensive amounts of mineralized tissue (black) within defects of the
autograft and rhBMP-7 group. In these groups solid defect bridging had
occurred. Notably, bone formation originated predominately from endosteal
areas as the periosteum - which is an additional source of mesenchymal
progenitor cells capable of differentiating into bone synthesizing osteoblasts -
had been removed during surgery. Untreated defects showed hardly any
new bone, defects treated with mPCL-TCP scaffolds considerably less bone
formation (generally at the scaffold-compact bone interface) and mainly soft
tissue was identified within the defect areas. Movat’s pentachrome staining
revealed that the soft tissue in the empty defects consisted mainly of fibrous
connective tissue. In contrast, the soft tissue in the scaffold only group
mainly resembled cartilage like tissue.
Bone formation in both the autograft and rhBMP-7 group occurred via
endochondral ossification, in which a cartilage template is gradually replaced
by a bone matrix. Deposited bone represented unorganized
cancellous/trabecular bone with mature osteocytes embedded in
characteristic lacunae. Bone in the autograft bone appeared to be less
mature when compared to bone in the rhBMP-7 group as higher amounts of
osteoid (red) and mineralized cartilage (green) were evident. Osteoid refers
to the unmineralized, organic portion of the bone matrix that forms prior to
the maturation of bone tissue (Fig. 23-25).
174
Fig. 23: Longitudinal native bone defect sections in the frontal plane (A-D)
embedded in PMMA and stained for Safranin O/von Kossa (E-H) and Movat’s
pentachrome (I-L) Von Kossa staining showed extensive amounts of mineralized
tissue (black) within defects of the autograft (F) and rhBMP-7 group (H) when
compared to untreated defects (E) or defects treated with mPCL-TCP scaffolds only
(G) where mainly soft tissue was identified (red). While the soft tissue in the empty
defects (I) consisted mainly of fibrous connective tissue (green), the soft tissue in
the scaffold only group (K) mainly resembled cartilage like tissue (light blue).
175
Fig. 24: PMMA sections stained for Movat’s pentachrome showing high amounts of
fibrous tissue (FT) in the untreated defects (A). In the scaffold only group (C) large
proportions of cartilaginous tissue (CT, light blue) were present while bone
formation (BT) was only observed along the scaffold/host bone interface evidenced
by osteoid islets (red, arrows). Bone in the autograft treated defects (B) appeared
less compact and of lower maturity when compare to the rhBMP-7 group (D) as
higher amounts of unmineralized osteoid and calcified cartilage (light green) were
found.
Fig. 25: High magnification images of the fibrous connective tissue (CT) in empty
defects with characteristic, elongated fibroblasts embedded in randomly aligned
collagenous fibre bundles (A). Maturing bone tissue (BT) of the autograft group (B)
and rhBMP-7 group (D) with unmineralized osteoid (red, arrows), mineralized
cartilage (light green), mineralized bone matrix (yellow), and mature osteocytes
embedded in characteristic lacunae (arrow heads). Image C shows connective
tissue (CT) and osteoblasts embedded in unmineralized osteoid (red, arrows) as
found in the scaffold only group (interface bone/scaffold).
176
Discussion
The present animal study was designed to investigate the effects of rhBMP-7
in combination with mPCL-TCP composite scaffolds built by computer aided
design using rapid prototyping technologies on bone healing. As a
prerequisite, a critical sized segmental bone defect model in ovine tibiae was
established showing a reliable non-union rate of 100%. Sheep as study
objects were chosen for their close analogies with human bone in terms of
remodelling and turnover as described in chapter I [126].
The transplantation of cancellous bone grafts represents the most frequently
chosen clinical treatment as these grafts possess osteoconductive and
osteoinductive properties. As a result, an experimental group in which
defects were treated with autologous bone graft from the iliac crest was
included as a positive control and benchmark.
A unilateral approach of the iliac crest to harvest graft was proven to be
sufficient to yield adequate amounts of graft for defect reconstruction. The
compressive strength of cancellous bone ranges between 2 and 20 MPa
depending on localization [250], however, the compressive strength of a
cancellous graft is considerably less (1-2 MPa) and therefore insufficient to
provide significant mechanical support [251]. Consequently, the required
stability to provide a loading environment facilitating bone healing has to be
provided by the osteosynthesis material. To minimize the risk of subsequent
implant failure, only partial weight bearing of the affected bone is hence
advisable until partial healing can be observed [4].
Bone graft substitutes that are capable of partially bearing loads and
disburden the osteosynthetic implants are therefore desirable and would
177
allow earlier load bearing whilst additionally stimulating the bone healing
process.
Bone graft substitutes for bone defect augmentation should consist of
materials that provide structural integrity and have a space-occupying effect
in order to maintain a proper reduction and continuity between the bone
fragments during surgery and to enable an adequate osteosynthesis to be
performed. Bioresorbable aliphatic polyesters, such as polyglycolide,
polylactide, polycaprolactone (PCL), and their copolymers, are suitable
materials for the design and fabrication of biocompatible scaffolds owing to
their significant track record for regulatory approval, and minimal
inflammatory and immunological responses evoked. As such, they have
been used as components in a plethora of devices for clinical applications
[252, 253]. These materials offer favourable surface chemistries for cell
attachment, proliferation, and differentiation, while degradation by-products
are nontoxic and metabolized/eliminated via natural pathways. Most
importantly, these thermoplastic polymers can easily be processed into
three-dimensional scaffolds with desired geometry, controlled porosity and
interconnectivity by applying modern computer-based solid free-form
fabrication methods [254]. These fabrication methods may therefore allow
the design of scaffolds with biomechanical properties similar to those of
cancellous bone [106].
For the present study, bioresorbable scaffolds of medical grade )-poly-
caprolactone incorporating 20% "-tricalcium phosphate (mPCL–TCP) were
produced by fused deposition modelling (FDM) as described previously
[169]. The structural parameters of the scaffolds were tailored by computer
178
aided design and included 100% pore interconnectivity within a range of
350–500 µm size, 70% scaffold porosity, and a 0/90° lay down pattern. This
architectural layout is particularly suitable for load bearing tissue engineering
applications since the fully interconnected network of scaffold fibres can
withstand early physiological and mechanical stress in a manner similar to
cancellous bone [169]. Moreover, the architectural pattern allows retaining of
coagulating blood during the early phase of healing, and bone in-growth at
later stages.
Importantly, scaffold handling during surgery was uncomplicated and
transplantation of these ready-to-use mPCL-TCP composites easily
achievable. As a result, economically, the transplantation of scaffolds
represents a time and therefore cost-effective alternative. Lohmann et al.
concluded that costs associated with the application of commercially
available bone replacement materials are comparable to those related to
bone grafts or may be even lower as high complication rates at the donor site
can significantly increase the total-costs-of-illness in the case of autografts
[255].
In animals where mPCL-TCP scaffolds were transplanted into the defects, no
signs of foreign body reactions to the transplant were observed underlining
their good biocompatibility. As evaluated by visual macroscopic assessment,
!CT analysis and histology, scaffolds were of structural integrity and had
integrated well at the host bone-scaffold interface. After 12 weeks no signs of
scaffold resorption were evident. Non-surprisingly, as biologically inactive,
cell-free materials, the scaffold type under investigation did not perform as
well as the gold standard autograft treatment as significantly lower values
179
were determined for torsional moment and stiffness as well as the
parameters describing new bone formation.
To introduce a biologically active, osteoinductive element, the scaffolds were
combined with the bone growth stimulating agent recombinant human bone
morphogenetic protein 7 (rhBMP-7) [73, 185]. Previous reports had shown
that rhBMP-7 is an efficient adjuvant in the healing of segmental bone
defects [4].
Bone morphogenetic proteins are a well-studied group of growth factors
involved in the processes of bone healing; and the human genome encodes
at least 20 of these multifunctional polypeptides [240]. However, only a
subset of BMPs, most notably BMP-2, -4, -7, and -9, have osteoinductive
activity [241]. The rhBMP-7 was provided in the form of Stryker®’s OP-1
implant which consists of 3.5 mg of rhBMP-7 (or Osteogenic Protein 1, OP-
1), formulated with 1 g purified bovine type I collagen carrier. The product is
commonly reconstituted with 2-3 cc of saline to form a paste which is then
implanted. Once implanted, rhBMP-7 stimulates natural bone healing by
actively recruiting stem cells from surrounding tissue and blood supply,
initiating the bone formation cascade [256]. OP-1 Implant is approved by the
FDA under a Humanitarian Device Exemption and is indicated for use as an
alternative to autograft in recalcitrant long bone non-unions where use of
autograft is unfeasible and alternative treatments have failed. Materials such
as the OP-1 implant do not provide any structural support as they are (semi-)
liquid materials. Therefore, the combination with a biocompatible scaffold of
suitable biomechanical properties appears appealing.
180
Importantly, rhBMP-7 should, however, not be used to treat patients who
have a known hypersensitivity to any of the components of the product
(about 18% of all patients develop antibodies to bovine type I collagen).
rhBMP-7 should not be applied at or near the vicinity of a resected tumour or
in patients with a history of malignancy and should also not be administered
to patients who are skeletally immature (<18 years of age or no radiographic
evidence of closure of epiphyses). As the potential effects of rhBMP-7
treatment on the human foetus have not been evaluated rhBMP-7 should not
be administered to pregnant women either. Studies in rats injected with high
doses of rhBMP-7 have shown that small amounts of rhBMP-7 will cross the
placental barrier. It has furthermore been reported that antibody formation to
rhBMP-7 occurs in up to 38% of patients the percentage of neutralizing
antibodies to rhBMP-7 was not determined separately. The clinical
significance of these antibodies, however, remains unclear [257].
Successful union requires rhBMP-7 to be retained at the surgical site long
enough to achieve osteoinduction. It is, however, unclear whether union
enhancement depends on the action of carrier-bound BMP or of freely
diffusible BMP slowly released from the carrier [258]. Nevertheless,
elucidation of retention kinetics may aid evaluation and development of BMP
delivery systems to optimize clinical outcomes. The mean residence time for
rhBMP-7 collagen putty was determined to be 10.4 ± 2.7 days in a spinal
fusion model (New Zealand White rabbits) [259]. These excretion profiles
and kinetic properties are similar to those described for rhBMP-2 (mean
residence times of 7.6 - 10.2 days) and include a biphasic release profile.
181
The initial burst release correlates with protein solubility, whereas the gradual
secondary release is governed by carrier properties [260, 261].
Regardless of the exact local pharmacokinetics, in the present study, the
release pattern of rhBMP-7 from the administered collagen particles and the
mPCL-TCP scaffold appeared to be consistent with the requirements of the
healing bone, as became evident by efficient healing of the defects in this
group. Biomechanical testing, histology, CT and !CT analysis consistently
showed comparable, in tendency even higher values and volumes of newly
formed bone in rhBMP-7 treated defects when compared to autograft
controls. Significant differences between rhBMP-7 and autograft group could,
however, not always be determined as the variation of bone formation within
the rhBMP-7 group was considerable which amongst other factors could be
due to variations of the local mechanical environment, in local pH, in
composition of the defect haematoma, surgical technique, release kinetics,
and the concentration of local connective tissue progenitor cells or degree of
vascularisation [262, 263].
All experimental studies utilizing rhBMP-7 as a bone growth stimulating
adjuvant illustrate a dose dependent effect and a species dependent
response variation [262, 264]. Maintaining a critical threshold concentration
of the rhBMP at the defect site for the necessary period of time (temporal
distribution) is crucial. Such supra-physiological dosages range from
0.01 mg/ml in small animal models such as rats to 0.4 mg/ml in rabbits to
more than 1.5 mg/ml in non-human primates. Different anatomical sites
require different therapeutic doses depending on the degree of
182
vascularisation, defect size and the number of resident responding cells.
Based on literature reports, in the present study, a standard application of
3.5 mg rhBMP-7 was chosen [265, 266]. The recommended dose of rhBMP-
7 for recalcitrant long bone non-unions in humans, however, is 7 mg or 2
vials (each containing 3.5 mg reconstituted with 1 g of type I bovine
collagen)[267].
Despite ample evidence of the benefit of BMPs in bone regeneration and
repair from preclinical and clinical studies, conclusive knowledge about BMP
dosage, time-course, release dynamics and carriers remains to be
determined. It is unclear why the impressive and convincing results seen in
vitro and in animal models are so far difficult to reproduce reliably in humans
[268]. Unfavourable release kinetics, insufficient mechanical scaffold stability
and porosity to allow cell and blood vessel infiltration into the carrier and
inflammatory tissue reactions might be few reasons. New delivery systems
with optimized and controlled release profiles may hence decrease or even
alleviate the need for excessive and expensive concentrations of BMPs.
As in the present study, segmental bone defects usually heal via indirect
repair mechanisms referred to as endochondral ossification. This process
involves the recruitment, proliferation, and differentiation of mesenchymal
progenitor cells into cartilage, which subsequently becomes calcified and
eventually is replaced by bone. The repair process is comprised of four
overlapping phases initiated by an immediate inflammatory response that
leads to the recruitment of mesenchymal progenitor cells and subsequent
differentiation into chondrocytes that produce cartilage and osteoblasts,
183
which form bone. After cartilage matrix is produced, it mineralizes, and a
transition from mineralized cartilage to bone occurs, initiated by the
resorption of mineralized cartilage [269]. It is the eventual bridging of hard
callus areas across the central defect gap that provides the initial
stabilization and regain of biomechanical function [270]. Primary bone
formation is followed by remodelling, in which the initial bony callus is
reshaped by secondary bone formation and resorption to restore the
anatomical structure that supports mechanical loads [271].
The current study results suggested not only a tendency towards increased
bone formation in defects treated with rhBMP-7 but also stimulated callus
maturation as evidenced by fewer amounts of osteoid and mineralized
cartilage in this group compared to autograft treated defects. These findings
could be attributed to rhBMP-7 accelerating the molecular events associated
with defect healing. Bone defect repair recapitulates the molecular pathways
of normal embryonic development with the coordinated participation of
several cell types originating from the cortex, periosteum, surrounding soft
tissue, and bone marrow space. The signalling molecules regulating this
process include pro-inflammatory cytokines, the transforming growth factor-
beta (TGF-") superfamily and other growth factors, and the angiogenic
factors [271, 272]. The transforming growth factor-beta (TGF-ß) superfamily
consists of a large number of growth and differentiation factors that include
bone morphogenetic proteins (BMPs), transforming growth factor beta (TGF-
ß), growth differentiation factors (GDFs), activins, inhibins, and Mullerian
inhibiting substance. Specific members of this family - such as BMPs (2-8),
GDF (1, 5, 8 and 10), and TGF-ß 1-3 - promote various stages of
184
intramembranous and endochondral bone ossification during bone healing
[273].
Recent studies demonstrate that BMP-7 specifically induces the expression
of numerous growth factors and multiple members of the BMP family
(including autoinduction), during the bone induction process [274, 275].
These factors then guide the bone formation process to completion.
Expression of BMP-7 was found to be enhanced in sites of endochondral
ossification about one week after fracture and it was therefore suggested that
BMP-7 acts predominately in the early stage of fracture healing [276]. In vitro
studies have shown that rhBMP-7 perpetuates mechanisms involved in bone
formation, including the promotion of chondrocyte maturation (extracellular
matrix component production) in normal articular chondrocytes, the
enhancement of osteoblastic characteristics (alkaline phosphate production)
of normal osteoblast cells, and the induction of the differentiation of
osteoclasts involved in the bone remodelling process. The end result of this
differentiation cascade is the production of weight-bearing bone with fully
functional bone marrow elements [277].
Considering the role of BMP-7 in mainly early stages of defect healing
together with the described release kinetics of BMP from collagen carriers,
and the favourable properties of the applied scaffolds, one could conclude
that in the present study, the requirements - based on the current knowledge
surrounding bone defect healing - have been met as closely as possible to
produce a suitable tissue engineered bone graft substitute.
185
Summary
In the current study, a standardized and reproducible ovine, tibial, critical-
sized defect model was developed as reflected by the non-union rate of
100% in the empty control group. The application of autografts from the iliac
crest caused defect bridging in all cases. When compared to intact controls,
the mechanical properties of the newly formed bone however were of inferior
quality after 3 months. Medical grade PCL-TCP scaffolds did not evoke
foreign body responses. The mPCL-TCP scaffolds alone did not result in any
substantial mineralization of the defect area. However, when combined with
rhBMP-7, bone formation within the defect equalled or even exceeded
amounts observed with autografts. The study results suggest that mPCL-
TCP scaffolds combined with a biologically active stimulus such as rhBMP-7
can serve as an equivalent alternative to autologous bone grafting in the
early phase of defect regeneration. These findings however must be
confirmed by long-term studies.
186
187
Conclusions and Recommendations
The detailed characterisation of cells involved in bone regeneration
associated processes is of utmost importance. Since sheep represent a
valuable model for human bone turnover and remodelling, data
characterising cells derived from these animals are of special interest. Ovine
marrow derived mesenchymal progenitor cells and cortical bone osteoblasts
exhibit morphological, immunophenotypical and multipotential characteristics
similar to those in humans, which underlines the value of sheep as a model
species. Unfortunately, available methods of analysis are limited as ovine
genes are only partially sequenced and very few antibodies specific for and
cross-reacting with equivalent sheep antigens are available. If further insight
into fundamental processes such as haematopoiesis, cell migration and
homing, injury repair, differentiation, and proliferation is to be provided, these
shortcomings need to be addressed in the future.
In vitro experiments identified mesenchymal progenitor cells reproducibly
displaying a higher osteogenic potential. In vivo, however, osteoblasts
exhibited a higher potential to form new bone. These results emphasize the
difficulties in extrapolating in vitro findings to in vivo settings and suggest that
osteoblasts isolated from compact bone possibly represent a suitable
alternative cell population for cell based tissue engineering applications.
When comparing the osteogenic potential of these cells after transplantation
in vivo, subcuntaneously transplanted cells showed a high degree of survival
and actively contribute to endochondral osteogenesis. Endochondral bone
188
formation, which describes a process in which a cartilage template is
gradually replaced by a bone matrix. It can also be observed in orthotopic
segmental defect sites. When compared to mesenchymal progenitor cells,
osteoblasts deposited higher amounts of new bone while osteoblast derived
bone was of higher maturation. Cell stimulation with rhBMP-7 increased the
rate of bone synthesis for both cell types and positively affected
neovascularisation and osteoclast activity. These results suggest that origin
and commitment of transplanted cells can determine type and degree of
ossification. They furthermore confirm that rhBMP-7 represents a potent
adjuvant stimulating bone formation.
It needs to be emphasized that microenvironmental conditions in ectopic
transplantation sites, again, may not be representative of specific cues cells
experience in a large segmental bone defect. However, it could also be
argued, that in such a defect cells are - similarly to ectopic sites - surrounded
by mainly soft tissue types rather than bone. Although an essential first step
was taken towards the characterization of ovine mesenchymal progenitor
cells and osteoblasts, essentially, further studies are required to verify these
findings in orthotopic models.
In the present study, the most important issues related to the establishment
of a large preclinical model for segmental bone defect research have been
addressed and discussed, and it was demonstrated how to develop such a
model. Regarding bench to bedside translations, an important milestone was
achieved in establishing a highly reproducible large animal model as an
189
essential prerequisite to systematically assess different bone grafting
materials, scaffolds, tissue engineered constructs and growth factors.
The performance of a novel tissue engineered construct was finally assessed
in a large animal model and compared to the standard autograft
transplantation. The application of autografts from the iliac crest caused
defect bridging in all cases. When compared to intact contralateral controls,
the mechanical properties of the newly formed bone, however, were of
inferior quality. Transplanted scaffolds showed good biocompatibility and did
not evoke foreign body responses. They did, however, not result in any
substantial mineralization of the defect area. However, when combined with
BMP-7, bone formation within the defects equalled or even exceeded
amounts observed with autografts. Therefore, the study results suggest that
these scaffolds combined with a biologically active stimulus such as BMP-7
can serve as an equivalent alternative to autologous bone grafting in the
early phase of defect regeneration. These findings however must be
confirmed by long-term studies (12-24 months) that also assess the events
surrounding scaffold degradation and bone remodelling. Lastly, if this novel
and promising technology is to be translated into a routine clinical application
with predictable outcomes stringent treatment indications/contraindications
need to be formulated and detailed dose-effect and dose-response
relationships determined taking into account variables such as defect size
and cause or patient age.
190
191
Bibliography
1. Perka C, Schultz O, Spitzer RS, Lindenhayn K, Burmester GR, Sittinger M. Segmental bone repair by tissue-engineered periosteal cell transplants with bioresorbable fleece and fibrin scaffolds in rabbits. Biomaterials 2000;21(11):1145-1153. 2. Komaki H, Tanaka T, Chazono M, Kikuchi T. Repair of segmental bone defects in rabbit tibiae using a complex of beta-tricalcium phosphate, type I collagen, and fibroblast growth factor-2. Biomaterials 2006;27(29):5118-5126. 3. Gugala Z, Gogolewski S. Healing of critical-size segmental bone defects in the sheep tibiae using bioresorbable polylactide membranes. Injury 2002;33 Suppl 2:B71-76. 4. den Boer FC, Wippermann BW, Blokhuis TJ, Patka P, Bakker FC, Haarman HJ. Healing of segmental bone defects with granular porous hydroxyapatite augmented with recombinant human osteogenic protein-1 or autologous bone marrow. J Orthop Res 2003;21(3):521-528. 5. Laurencin C, Khan Y, El-Amin SF. Bone graft substitutes. Expert review of medical devices 2006;3(1):49-57. 6. Wildemann B, Kadow-Romacker A, Pruss A, Haas NP, Schmidmaier G. Quantification of growth factors in allogenic bone grafts extracted with three different methods. Cell and tissue banking 2007;8(2):107-114. 7. Perry CR. Bone repair techniques, bone graft, and bone graft substitutes. Clin Orthop Relat Res 1999(360):71-86. 8. DeCoster TA, Gehlert RJ, Mikola EA, Pirela-Cruz MA. Management of posttraumatic segmental bone defects. J Am Acad Orthop Surg 2004;12(1):28-38. 9. Clements JR, Carpenter BB, Pourciau JK. Treating segmental bone defects: a new technique. J Foot Ankle Surg 2008;47(4):350-356. 10. Lane JM, Tomin E, Bostrom MP. Biosynthetic bone grafting. Clin Orthop Relat Res 1999(367 Suppl):S107-117. 11. Einhorn TA, Lane JM, Burstein AH, Kopman CR, Vigorita VJ. The healing of segmental bone defects induced by demineralized bone matrix. A radiographic and biomechanical study. J Bone Joint Surg Am 1984;66(2):274-279. 12. Liu G, Zhao L, Zhang W, Cui L, Liu W, Cao Y. Repair of goat tibial defects with bone marrow stromal cells and beta-tricalcium phosphate. J Mater Sci Mater Med 2008;19(6):2367-2376.
192
13. Blokhuis TJ, Wippermann BW, den Boer FC, van Lingen A, Patka P, Bakker FC, et al. Resorbable calcium phosphate particles as a carrier material for bone marrow in an ovine segmental defect. J Biomed Mater Res 2000;51(3):369-375. 14. Gao TJ, Lindholm TS, Kommonen B, Ragni P, Paronzini A, Lindholm TC, et al. Enhanced healing of segmental tibial defects in sheep by a composite bone substitute composed of tricalcium phosphate cylinder, bone morphogenetic protein, and type IV collagen. J Biomed Mater Res 1996;32(4):505-512. 15. Theos C, Koulouvaris P, Kottakis S, Demertzis N. Reconstruction of tibia defects by ipsilateral vascularized fibula transposition. Arch Orthop Trauma Surg 2008;128(2):179-184. 16. Sheller MR, Crowther RS, Kinney JH, Yang J, Di Jorio S, Breunig T, et al. Repair of rabbit segmental defects with the thrombin peptide, TP508. J Orthop Res 2004;22(5):1094-1099. 17. Bucholz RW, Carlton A, Holmes R. Interporous hydroxyapatite as a bone graft substitute in tibial plateau fractures. Clin Orthop Relat Res 1989(240):53-62. 18. Oest ME, Dupont KM, Kong HJ, Mooney DJ, Guldberg RE. Quantitative assessment of scaffold and growth factor-mediated repair of critically sized bone defects. J Orthop Res 2007;25(7):941-950. 19. Stevenson S. Enhancement of fracture healing with autogenous and allogeneic bone grafts. Clin Orthop Relat Res 1998(355 Suppl):S239-246. 20. Sciadini MF, Dawson JM, Johnson KD. Evaluation of bovine-derived bone protein with a natural coral carrier as a bone-graft substitute in a canine segmental defect model. J Orthop Res 1997;15(6):844-857. 21. Younger EM, Chapman MW. Morbidity at bone graft donor sites. J Orthop Trauma 1989;3(3):192-195. 22. Taylor GI, Miller GD, Ham FJ. The free vascularized bone graft. A clinical extension of microvascular techniques. Plast Reconstr Surg 1975;55(5):533-544. 23. Dell PC, Burchardt H, Glowczewskie FP, Jr. A roentgenographic, biomechanical, and histological evaluation of vascularized and non-vascularized segmental fibular canine autografts. J Bone Joint Surg Am 1985;67(1):105-112. 24. Lindsey RW, Gugala Z, Milne E, Sun M, Gannon FH, Latta LL. The efficacy of cylindrical titanium mesh cage for the reconstruction of a critical-
193
size canine segmental femoral diaphyseal defect. J Orthop Res 2006;24(7):1438-1453. 25. Puelacher WC, Vacanti JP, Ferraro NF, Schloo B, Vacanti CA. Femoral shaft reconstruction using tissue-engineered growth of bone. Int J Oral Maxillofac Surg 1996;25(3):223-228. 26. Gazdag AR, Lane JM, Glaser D, Forster RA. Alternatives to Autogenous Bone Graft: Efficacy and Indications. J Am Acad Orthop Surg 1995;3(1):1-8. 27. Muscolo DL, Ayerza MA, Aponte-Tinao LA. Massive allograft use in orthopedic oncology. Orthop Clin North Am 2006;37(1):65-74. 28. Chapman MW, Bucholz R, Cornell C. Treatment of acute fractures with a collagen-calcium phosphate graft material. A randomized clinical trial. J Bone Joint Surg Am 1997;79(4):495-502. 29. Cacchioli A, Spaggiari B, Ravanetti F, Martini FM, Borghetti P, Gabbi C. The critical sized bone defect: morphological study of bone healing. Ann Fac Medic Vet di Parma 2006;26:97-110. 30. Cierny G, 3rd, Zorn KE. Segmental tibial defects. Comparing conventional and Ilizarov methodologies. Clin Orthop Relat Res 1994(301):118-123. 31. Ilizarov GA. The tension-stress effect on the genesis and growth of tissues: Part II. The influence of the rate and frequency of distraction. Clin Orthop Relat Res 1989(239):263-285. 32. Goldstrohm GL, Mears DC, Swartz WM. The results of 39 fractures complicated by major segmental bone loss and/or leg length discrepancy. The Journal of trauma 1984;24(1):50-58. 33. Gugala Z, Lindsey RW, Gogolewski S. New Approaches in the Treatment of Critical-Size Segmental Defects in Long Bones. Macromol Symp 2007;253:147–161. 34. O'Loughlin PF, Morr S, Bogunovic L, Kim AD, Park B, Lane JM. Selection and development of preclinical models in fracture-healing research. J Bone Joint Surg Am 2008;90 Suppl 1:79-84. 35. Einhorn TA. Clinically applied models of bone regeneration in tissue engineering research. Clin Orthop Relat Res 1999(367 Suppl):S59-67. 36. Rimondini L, Nicoli-Aldini N, Fini M, Guzzardella G, Tschon M, Giardino R. In vivo experimental study on bone regeneration in critical bone defects using an injectable biodegradable PLA/PGA copolymer. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 2005;99(2):148-154.
194
37. Gugala Z, Gogolewski S. Regeneration of segmental diaphyseal defects in sheep tibiae using resorbable polymeric membranes: a preliminary study. J Orthop Trauma 1999;13(3):187-195. 38. Tseng SS, Lee MA, Reddi AH. Nonunions and the potential of stem cells in fracture-healing. J Bone Joint Surg Am 2008;90 Suppl 1:92-98. 39. Egermann M, Goldhahn J, Schneider E. Animal models for fracture treatment in osteoporosis. Osteoporos Int 2005;16 Suppl 2:S129-138. 40. Liebschner MA. Biomechanical considerations of animal models used in tissue engineering of bone. Biomaterials 2004;25(9):1697-1714. 41. Schimandle JH, Boden SD. Spine update. The use of animal models to study spinal fusion. Spine 1994;19(17):1998-2006. 42. Pearce AI, Richards RG, Milz S, Schneider E, Pearce SG. Animal models for implant biomaterial research in bone: a review. Eur Cell Mater 2007;13:1-10. 43. Martini L, Fini M, Giavaresi G, Giardino R. Sheep model in orthopedic research: a literature review. Comparative medicine 2001;51(4):292-299. 44. Aerssens J, Boonen S, Lowet G, Dequeker J. Interspecies differences in bone composition, density, and quality: potential implications for in vivo bone research. Endocrinology 1998;139(2):663-670. 45. Gong JK, Arnold JS, Cohn SH. The Density of Organic and Volatile and Non-Volatile Inorganic Components of Bone. The Anatomical record 1964;149:319-324. 46. Wang X, Mabrey JD, Agrawal CM. An interspecies comparison of bone fracture properties. Bio-medical materials and engineering 1998;8(1):1-9. 47. Bloebaum RD, Ota DT, Skedros JG, Mantas JP. Comparison of human and canine external femoral morphologies in the context of total hip replacement. J Biomed Mater Res 1993;27(9):1149-1159. 48. Kaneps AJ, Stover SM, Lane NE. Changes in canine cortical and cancellous bone mechanical properties following immobilization and remobilization with exercise. Bone 1997;21(5):419-423. 49. Mittra EQ, Y-X. MATERIAL PROPERTIES OF SHEEP TRABECULAR BONE DETERMINED BY NANOMECHANICAL TESTING. IEEE 2003:229-230. 50. Mittra E, Rubin C, Qin YX. Interrelationship of trabecular mechanical and microstructural properties in sheep trabecular bone. J Biomech 2005;38(6):1229-1237.
195
51. Nafei A, Danielsen CC, Linde F, Hvid I. Properties of growing trabecular ovine bone. Part I: mechanical and physical properties. J Bone Joint Surg Br 2000;82(6):910-920. 52. Makoto S, Kenji S, Koichi K, Jun S, Eisuke T, Toshiaki H. Measurements of Mechanical Properties of Cortical Bone Using Nanoindentation Tests. Proceedings of Annual Meeting of Japanese Society for Orthopaedic Biomechanics; 2003; 2003. p. 1-5. 53. Neyt JG, Buckwalter JA, Carroll NC. Use of animal models in musculoskeletal research. The Iowa orthopaedic journal 1998;18:118-123. 54. Newman E, Turner AS, Wark JD. The potential of sheep for the study of osteopenia: current status and comparison with other animal models. Bone 1995;16(4 Suppl):277S-284S. 55. Taylor WR, Ehrig RM, Duda GN, Schell H, Seebeck P, Heller MO. On the influence of soft tissue coverage in the determination of bone kinematics using skin markers. J Orthop Res 2005;23(4):726-734. 56. Taylor WR, Ehrig RM, Heller MO, Schell H, Seebeck P, Duda GN. Tibio-femoral joint contact forces in sheep. J Biomech 2006;39(5):791-798. 57. Ravaglioli A, Krajewski A, Celotti GC, Piancastelli A, Bacchini B, Montanari L, et al. Mineral evolution of bone. Biomaterials 1996;17(6):617-622. 58. Anderson ML, Dhert WJ, de Bruijn JD, Dalmeijer RA, Leenders H, van Blitterswijk CA, et al. Critical size defect in the goat's os ilium. A model to evaluate bone grafts and substitutes. Clin Orthop Relat Res 1999(364):231-239. 59. den Boer FC, Patka P, Bakker FC, Wippermann BW, van Lingen A, Vink GQ, et al. New segmental long bone defect model in sheep: quantitative analysis of healing with dual energy x-ray absorptiometry. J Orthop Res 1999;17(5):654-660. 60. de Kleer V. Development of bone. In: Sumner-Smith G, editor. Bone in Clinical Orthopedics. Philadelphia: W. B. Saunders Co, 2006. p. 1-80. 61. Eitel F, Klapp F, Jacobson W, Schweiberer L. Bone regeneration in animals and in man. A contribution to understanding the relative value of animal experiments to human pathophysiology. Arch Orthop Trauma Surg 1981;99(1):59-64. 62. Dai KR, Xu XL, Tang TT, Zhu ZA, Yu CF, Lou JR, et al. Repairing of goat tibial bone defects with BMP-2 gene-modified tissue-engineered bone. Calcified tissue international 2005;77(1):55-61.
196
63. Mosekilde L, Weisbrode SE, Safron JA, Stills HF, Jankowsky ML, Ebert DC, et al. Calcium-restricted ovariectomized Sinclair S-1 minipigs: an animal model of osteopenia and trabecular plate perforation. Bone 1993;14(3):379-382. 64. Thorwarth M, Schultze-Mosgau S, Kessler P, Wiltfang J, Schlegel KA. Bone regeneration in osseous defects using a resorbable nanoparticular hydroxyapatite. J Oral Maxillofac Surg 2005;63(11):1626-1633. 65. Macdonald W, Skirving AP, Scull ER. A device for producing experimental fractures. Acta Orthop Scand 1988;59(5):542-544. 66. Skirving AP, Day R, Macdonald W, McLaren R. Carbon fiber reinforced plastic (CFRP) plates versus stainless steel dynamic compression plates in the treatment of fractures of the tibiae in dogs. Clin Orthop Relat Res 1987(224):117-124. 67. Schemitsch EH, Kowalski MJ, Swiontkowski MF. Soft-tissue blood flow following reamed versus unreamed locked intramedullary nailing: a fractured sheep tibia model. Annals of plastic surgery 1996;36(1):70-75. 68. Schemitsch EH, Kowalski MJ, Swiontkowski MF, Senft D. Cortical bone blood flow in reamed and unreamed locked intramedullary nailing: a fractured tibia model in sheep. J Orthop Trauma 1994;8(5):373-382. 69. Tepic S, Remiger AR, Morikawa K, Predieri M, Perren SM. Strength recovery in fractured sheep tibia treated with a plate or an internal fixator: an experimental study with a two-year follow-up. J Orthop Trauma 1997;11(1):14-23. 70. Gilbert JA, Dahners LE, Atkinson MA. The effect of external fixation stiffness on early healing of transverse osteotomies. J Orthop Res 1989;7(3):389-397. 71. Tiedeman JJ, Lippiello L, Connolly JF, Strates BS. Quantitative roentgenographic densitometry for assessing fracture healing. Clin Orthop Relat Res 1990(253):279-286. 72. Markel MD, Wikenheiser MA, Chao EY. A study of fracture callus material properties: relationship to the torsional strength of bone. J Orthop Res 1990;8(6):843-850. 73. Faria ML, Lu Y, Heaney K, Uthamanthil RK, Muir P, Markel MD. Recombinant human bone morphogenetic protein-2 in absorbable collagen sponge enhances bone healing of tibial osteotomies in dogs. Vet Surg 2007;36(2):122-131. 74. Edwards RB, 3rd, Seeherman HJ, Bogdanske JJ, Devitt J, Vanderby R, Jr., Markel MD. Percutaneous injection of recombinant human bone
197
morphogenetic protein-2 in a calcium phosphate paste accelerates healing of a canine tibial osteotomy. J Bone Joint Surg Am 2004;86-A(7):1425-1438. 75. Wang CJ, Huang HY, Chen HH, Pai CH, Yang KD. Effect of shock wave therapy on acute fractures of the tibia: a study in a dog model. Clin Orthop Relat Res 2001(387):112-118. 76. Hupel TM, Weinberg JA, Aksenov SA, Schemitsch EH. Effect of unreamed, limited reamed, and standard reamed intramedullary nailing on cortical bone porosity and new bone formation. J Orthop Trauma 2001;15(1):18-27. 77. Jain R, Podworny N, Hupel TM, Weinberg J, Schemitsch EH. Influence of plate design on cortical bone perfusion and fracture healing in canine segmental tibial fractures. J Orthop Trauma 1999;13(3):178-186. 78. Nakamura T, Hara Y, Tagawa M, Tamura M, Yuge T, Fukuda H, et al. Recombinant human basic fibroblast growth factor accelerates fracture healing by enhancing callus remodeling in experimental dog tibial fracture. J Bone Miner Res 1998;13(6):942-949. 79. Heitemeyer U, Claes L, Hierholzer G. [The significance of postoperative stability for osseous repair of a multiple fragment fracture. Animal experiment studies]. Unfallchirurg 1990;93(2):49-55. 80. Goodship AE, Watkins PE, Rigby HS, Kenwright J. The role of fixator frame stiffness in the control of fracture healing. An experimental study. J Biomech 1993;26(9):1027-1035. 81. Wallace AL, Makki R, Weiss JB, Hughes SP. Measurement of serum angiogenic factor in devascularized experimental tibial fractures. J Orthop Trauma 1995;9(4):324-332. 82. Augat P, Burger J, Schorlemmer S, Henke T, Peraus M, Claes L. Shear movement at the fracture site delays healing in a diaphyseal fracture model. J Orthop Res 2003;21(6):1011-1017. 83. Augat P, Merk J, Genant HK, Claes L. Quantitative assessment of experimental fracture repair by peripheral computed tomography. Calcified tissue international 1997;60(2):194-199. 84. Epari DR, Schell H, Bail HJ, Duda GN. Instability prolongs the chondral phase during bone healing in sheep. Bone 2006;38(6):864-870. 85. Schell H, Epari DR, Kassi JP, Bragulla H, Bail HJ, Duda GN. The course of bone healing is influenced by the initial shear fixation stability. J Orthop Res 2005;23(5):1022-1028. 86. Hente R, Cordey J, Rahn BA, Maghsudi M, von Gumppenberg S, Perren SM. Fracture healing of the sheep tibia treated using a unilateral
198
external fixator. Comparison of static and dynamic fixation. Injury 1999;30 Suppl 1:A44-51. 87. Hantes ME, Mavrodontidis AN, Zalavras CG, Karantanas AH, Karachalios T, Malizos KN. Low-intensity transosseous ultrasound accelerates osteotomy healing in a sheep fracture model. J Bone Joint Surg Am 2004;86-A(10):2275-2282. 88. Epari DR, Lienau J, Schell H, Witt F, Duda GN. Pressure, oxygen tension and temperature in the periosteal callus during bone healing-An in vivo study in sheep. Bone 2008. 89. Claes L, Augat P, Schorlemmer S, Konrads C, Ignatius A, Ehrnthaller C. Temporary distraction and compression of a diaphyseal osteotomy accelerates bone healing. J Orthop Res 2008;26(6):772-777. 90. Kuttenberger JJ, Stubinger S, Waibel A, Werner M, Klasing M, Ivanenko M, et al. Computer-guided CO2-laser osteotomy of the sheep tibia: technical prerequisites and first results. Photomedicine and laser surgery 2008;26(2):129-136. 91. Gao TJ, Lindholm TS, Kommonen B, Ragni P, Paronzini A, Lindholm TC, et al. The use of a coral composite implant containing bone morphogenetic protein to repair a segmental tibial defect in sheep. 1997;21(3):194-200. 92. Wefer J, Wefer A, Schratt HE, Thermann H, Wippermann BW. [Healing of autologous cancellous bone transplants and hydroxylapatite ceramics in tibial segment defects. Value of ultrasonic follow up]. Unfallchirurg 2000;103(6):452-461. 93. Bloemers FW, Blokhuis TJ, Patka P, Bakker FC, Wippermann BW, Haarman HJ. Autologous bone versus calcium-phosphate ceramics in treatment of experimental bone defects. Journal of biomedical materials research 2003;66(2):526-531. 94. Meinel L, Zoidis E, Zapf J, Hassa P, Hottiger MO, Auer JA, et al. Localized insulin-like growth factor I delivery to enhance new bone formation. Bone 2003;33(4):660-672. 95. Mastrogiacomo M, Corsi A, Francioso E, Di Comite M, Monetti F, Scaglione S, et al. Reconstruction of extensive long bone defects in sheep using resorbable bioceramics based on silicon stabilized tricalcium phosphate. Tissue Eng 2006;12(5):1261-1273. 96. Maissen O, Eckhardt C, Gogolewski S, Glatt M, Arvinte T, Steiner A, et al. Mechanical and radiological assessment of the influence of rhTGFbeta-3 on bone regeneration in a segmental defect in the ovine tibia: pilot study. J Orthop Res 2006;24(8):1670-1678.
199
97. Sarkar MR, Augat P, Shefelbine SJ, Schorlemmer S, Huber-Lang M, Claes L, et al. Bone formation in a long bone defect model using a platelet-rich plasma-loaded collagen scaffold. Biomaterials 2006;27(9):1817-1823. 98. Tyllianakis M, Deligianni D, Panagopoulos A, Pappas M, Sourgiadaki E, Mavrilas D, et al. Biomechanical comparison of callus over a locked intramedullary nail in various segmental bone defects in a sheep model. Med Sci Monit 2007;13(5):BR125-130. 99. Teixeira CR, Rahal SC, Volpi RS, Taga R, Cestari TM, Granjeiro JM, et al. Tibial segmental bone defect treated with bone plate and cage filled with either xenogeneic composite or autologous cortical bone graft. An experimental study in sheep. Vet Comp Orthop Traumatol 2007;20(4):269-276. 100. Schneiders W, Reinstorf A, Biewener A, Serra A, Grass R, Kinscher M, et al. In vivo effects of modification of hydroxyapatite/collagen composites with and without chondroitin sulphate on bone remodeling in the sheep tibia. J Orthop Res 2008. 101. Rozen N, Bick T, Bajayo A, Shamian B, Schrift-Tzadok M, Gabet Y, et al. Transplanted blood-derived endothelial progenitor cells (EPC) enhance bridging of sheep tibia critical size defects. Bone 2009. 102. Niemeyer P, Schonberger TS, Hahn J, Kasten P, Fellenberg J, Suedkamp N, et al. Xenogenic transplantation of human mesenchymal stem cells in a critical size defect of the sheep tibia for bone regeneration. Tissue Eng Part A 2009. 103. Claes L, Eckert-Hubner K, Augat P. The fracture gap size influences the local vascularization and tissue differentiation in callus healing. Langenbecks Arch Surg 2003;388(5):316-322. 104. Duda GN, Sollmann M, Sporrer S, Hoffmann JE, Kassi JP, Khodadadyan C, et al. Interfragmentary motion in tibial osteotomies stabilized with ring fixators. Clin Orthop Relat Res 2002(396):163-172. 105. Kaspar K, Matziolis G, Strube P, Senturk U, Dormann S, Bail HJ, et al. A new animal model for bone atrophic nonunion: Fixation by external fixator. J Orthop Res 2008. 106. Hutmacher DW, Schantz JT, Lam CX, Tan KC, Lim TC. State of the art and future directions of scaffold-based bone engineering from a biomaterials perspective. J Tissue Eng Regen Med 2007;1(4):245-260. 107. Kettunen J, Makela A, Miettinen H, Nevalainen T, Heikkila M, Tormala P, et al. Fixation of femoral shaft osteotomy with an intramedullary composite rod: an experimental study on dogs with a two-year follow-up. J Biomater Sci Polym Ed 1999;10(1):33-45.
200
108. Sinicropi SM, Su BW, Raia FJ, Parisien M, Strauch RJ, Rosenwasser MP. The effects of implant composition on extensor tenosynovitis in a canine distal radius fracture model. J Hand Surg [Am] 2005;30(2):300-307. 109. An YH, Friedman RJ, Powers DL, Draughn RA, Latour RA, Jr. Fixation of osteotomies using bioabsorbable screws in the canine femur. Clin Orthop Relat Res 1998(355):300-311. 110. Aro HT, Chao EY. Bone-healing patterns affected by loading, fracture fragment stability, fracture type, and fracture site compression. Clin Orthop Relat Res 1993(293):8-17. 111. Peter CP, Cook WO, Nunamaker DM, Provost MT, Seedor JG, Rodan GA. Effect of alendronate on fracture healing and bone remodeling in dogs. J Orthop Res 1996;14(1):74-79. 112. Wolf S, Janousek A, Pfeil J, Veith W, Haas F, Duda G, et al. The effects of external mechanical stimulation on the healing of diaphyseal osteotomies fixed by flexible external fixation. Clin Biomech (Bristol, Avon) 1998;13(4-5):359-364. 113. Schell H, Thompson MS, Bail HJ, Hoffmann JE, Schill A, Duda GN, et al. Mechanical induction of critically delayed bone healing in sheep: Radiological and biomechanical results. J Biomech 2008. 114. Curtis MJ, Brown PR, Dick JD, Jinnah RH. Contaminated fractures of the tibia: a comparison of treatment modalities in an animal model. J Orthop Res 1995;13(2):286-295. 115. den Boer FC, Bramer JA, Blokhuis TJ, Van Soest EJ, Jenner JM, Patka P, et al. Effect of recombinant human osteogenic protein-1 on the healing of a freshly closed diaphyseal fracture. Bone 2002;31(1):158-164. 116. Krappinger D, Schubert H, Wenzel V, Rieger M, Stadlbauer KH, Blauth M, et al. A pelvic fracture model for the assessment of treatment options in a laboratory environment. Injury 2007;38(10):1151-1157. 117. Kleinman PL, Zurakowski D, Strauss KJ, Cleveland RH, Perez-Rosello JM, Nichols DP, et al. Detection of simulated inflicted metaphyseal fractures in a fetal pig model: image optimization and dose reduction with computed radiography. Radiology 2008;247(2):381-390. 118. Hill PF, Watkins PE. The prevention of experimental osteomyelitis in a model of gunshot fracture in the pig. Eur J Orthop Surg Traumatol 2001;11:237-241. 119. Bail HJ, Kolbeck S, Krummrey G, Schmidmaier G, Haas NP, Raschke MJ. Systemic application of growth hormone for enhancement of secondary and intramembranous fracture healing. Hormone research 2002;58 Suppl 3:39-42.
201
120. Allan DG, Russell GG, Moreau MJ, Raso VJ, Budney D. Vertebral end-plate failure in porcine and bovine models of spinal fracture instrumentation. J Orthop Res 1990;8(1):154-156. 121. Russell G, Lavoie G, Evenson R, Moreau M, Budney D, Raso VJ. A reproducible porcine vertebral fracture for biomechanical testing of spinal fixation devices. Clin Orthop Relat Res 1992(284):267-272. 122. Hillier ML, Bell LS. Differentiating human bone from animal bone: a review of histological methods. J Forensic Sci 2007;52(2):249-263. 123. Turner AS. The sheep as a model for osteoporosis in humans. Vet J 2002;163(3):232-239. 124. Dyce K.M. SWO, Wensing C.J.G. Textbook of veterinary Anatomy. 3rd ed. Philadelphia: Saunders 2002. 125. Getty R. Sissons and Grossman's The Anatomy of the Domestic Animals. 5th ed. Philadelphia: W B Saunders, 1975. 126. Reichert JC, Saifzadeh S, Wullschleger ME, Epari DR, Schutz MA, Duda GN, et al. The challenge of establishing preclinical models for segmental bone defect research. Biomaterials 2009;30(12):2149-2163. 127. Gronthos S, Mrozik K, Shi S, Bartold PM. Ovine periodontal ligament stem cells: isolation, characterization, and differentiation potential. Calcified tissue international 2006;79(5):310-317. 128. McCarty RC, Gronthos S, Zannettino AC, Foster BK, Xian CJ. Characterisation and developmental potential of ovine bone marrow derived mesenchymal stem cells. Journal of cellular physiology 2008;219(2):324-333. 129. Sheehan DC, Hrapchak BB. Theory and practice of histotechnology. 2 ed. St. Louis: Mosby, C. V., 1980. 130. Gronthos S, Mankani M, Brahim J, Robey PG, Shi S. Postnatal human dental pulp stem cells (DPSCs) in vitro and in vivo. Proceedings of the National Academy of Sciences of the United States of America 2000;97(25):13625-13630. 131. Gronthos S, Franklin DM, Leddy HA, Robey PG, Storms RW, Gimble JM. Surface protein characterization of human adipose tissue-derived stromal cells. Journal of cellular physiology 2001;189(1):54-63. 132. Pittenger MF, Mackay AM, Beck SC, Jaiswal RK, Douglas R, Mosca JD, et al. Multilineage potential of adult human mesenchymal stem cells. Science (New York, NY 1999;284(5411):143-147.
202
133. Shin H, Zygourakis K, Farach-Carson MC, Yaszemski MJ, Mikos AG. Modulation of differentiation and mineralization of marrow stromal cells cultured on biomimetic hydrogels modified with Arg-Gly-Asp containing peptides. J Biomed Mater Res A 2004;69(3):535-543. 134. Porter JR, Henson A, Popat KC. Biodegradable poly(epsilon-caprolactone) nanowires for bone tissue engineering applications. Biomaterials 2008. 135. Martin I, Wendt D, Heberer M. The role of bioreactors in tissue engineering. Trends in biotechnology 2004;22(2):80-86. 136. Castro-Malaspina H, Gay RE, Resnick G, Kapoor N, Meyers P, Chiarieri D, et al. Characterization of human bone marrow fibroblast colony-forming cells (CFU-F) and their progeny. Blood 1980;56(2):289-301. 137. Friedenstein AJ. Stromal mechanisms of bone marrow: cloning in vitro and retransplantation in vivo. Haematology and blood transfusion 1980;25:19-29. 138. Owen M. Marrow stromal stem cells. Journal of cell science 1988;10:63-76. 139. Gronthos S, Zannettino AC, Hay SJ, Shi S, Graves SE, Kortesidis A, et al. Molecular and cellular characterisation of highly purified stromal stem cells derived from human bone marrow. Journal of cell science 2003;116(Pt 9):1827-1835. 140. Simmons PJ, Torok-Storb B. Identification of stromal cell precursors in human bone marrow by a novel monoclonal antibody, STRO-1. Blood 1991;78(1):55-62. 141. Kadiyala S, Young RG, Thiede MA, Bruder SP. Culture expanded canine mesenchymal stem cells possess osteochondrogenic potential in vivo and in vitro. Cell transplantation 1997;6(2):125-134. 142. Martin DR, Cox NR, Hathcock TL, Niemeyer GP, Baker HJ. Isolation and characterization of multipotential mesenchymal stem cells from feline bone marrow. Experimental hematology 2002;30(8):879-886. 143. Yoshimura H, Muneta T, Nimura A, Yokoyama A, Koga H, Sekiya I. Comparison of rat mesenchymal stem cells derived from bone marrow, synovium, periosteum, adipose tissue, and muscle. Cell and tissue research 2007;327(3):449-462. 144. Gronthos S, Brahim J, Li W, Fisher LW, Cherman N, Boyde A, et al. Stem cell properties of human dental pulp stem cells. Journal of dental research 2002;81(8):531-535.
203
145. Hayflick L. The Limited in Vitro Lifetime of Human Diploid Cell Strains. Experimental cell research 1965;37:614-636. 146. Rheinwald JG, Green H. Serial cultivation of strains of human epidermal keratinocytes: the formation of keratinizing colonies from single cells. Cell 1975;6(3):331-343. 147. Thornton SC, Mueller SN, Levine EM. Human endothelial cells: use of heparin in cloning and long-term serial cultivation. Science (New York, NY 1983;222(4624):623-625. 148. Bruder SP, Jaiswal N, Haynesworth SE. Growth kinetics, self-renewal, and the osteogenic potential of purified human mesenchymal stem cells during extensive subcultivation and following cryopreservation. Journal of cellular biochemistry 1997;64(2):278-294. 149. Digirolamo CM, Stokes D, Colter D, Phinney DG, Class R, Prockop DJ. Propagation and senescence of human marrow stromal cells in culture: a simple colony-forming assay identifies samples with the greatest potential to propagate and differentiate. British journal of haematology 1999;107(2):275-281. 150. Stanford CM, Jacobson PA, Eanes ED, Lembke LA, Midura RJ. Rapidly forming apatitic mineral in an osteoblastic cell line (UMR 106-01 BSP). The Journal of biological chemistry 1995;270(16):9420-9428. 151. Gregory CA, Gunn WG, Peister A, Prockop DJ. An Alizarin red-based assay of mineralization by adherent cells in culture: comparison with cetylpyridinium chloride extraction. Analytical biochemistry 2004;329(1):77-84. 152. Niyibizi C, Eyre DR. Structural characteristics of cross-linking sites in type V collagen of bone. Chain specificities and heterotypic links to type I collagen. European journal of biochemistry / FEBS 1994;224(3):943-950. 153. Viguet-Carrin S, Garnero P, Delmas PD. The role of collagen in bone strength. Osteoporos Int 2006;17(3):319-336. 154. Price PA, Otsuka AA, Poser JW, Kristaponis J, Raman N. Characterization of a gamma-carboxyglutamic acid-containing protein from bone. Proceedings of the National Academy of Sciences of the United States of America 1976;73(5):1447-1451. 155. Boskey AL, Gadaleta S, Gundberg C, Doty SB, Ducy P, Karsenty G. Fourier transform infrared microspectroscopic analysis of bones of osteocalcin-deficient mice provides insight into the function of osteocalcin. Bone 1998;23(3):187-196. 156. Choi ST, Kim JH, Kang EJ, Lee SW, Park MC, Park YB, et al. Osteopontin might be involved in bone remodelling rather than in
204
inflammation in ankylosing spondylitis. Rheumatology (Oxford, England) 2008;47(12):1775-1779. 157. Wagner HD, Weiner S. On the relationship between the microstructure of bone and its mechanical stiffness. J Biomech 1992;25(11):1311-1320. 158. Datta N, Pham QP, Sharma U, Sikavitsas VI, Jansen JA, Mikos AG. In vitro generated extracellular matrix and fluid shear stress synergistically enhance 3D osteoblastic differentiation. Proceedings of the National Academy of Sciences of the United States of America 2006;103(8):2488-2493. 159. Knothe Tate ML, Falls TD, McBride SH, Atit R, Knothe UR. Mechanical modulation of osteochondroprogenitor cell fate. The international journal of biochemistry & cell biology 2008;40(12):2720-2738. 160. Zhou Y, Chen F, Ho ST, Woodruff MA, Lim TM, Hutmacher DW. Combined marrow stromal cell-sheet techniques and high-strength biodegradable composite scaffolds for engineered functional bone grafts. Biomaterials 2007;28(5):814-824. 161. Abdallah BM, Kassem M. The use of mesenchymal (skeletal) stem cells for treatment of degenerative diseases: current status and future perspectives. Journal of cellular physiology 2009;218(1):9-12. 162. Griffith LG, Swartz MA. Capturing complex 3D tissue physiology in vitro. Nature reviews 2006;7(3):211-224. 163. Hutmacher DW, Cool S. Concepts of scaffold-based tissue engineering--the rationale to use solid free-form fabrication techniques. Journal of cellular and molecular medicine 2007;11(4):654-669. 164. Arthur A, Zannettino A, Gronthos S. The therapeutic applications of multipotential mesenchymal/stromal stem cells in skeletal tissue repair. Journal of cellular physiology 2009;218(2):237-245. 165. Biggs MJ, Richards RG, Gadegaard N, Wilkinson CD, Oreffo RO, Dalby MJ. The use of nanoscale topography to modulate the dynamics of adhesion formation in primary osteoblasts and ERK/MAPK signalling in STRO-1+ enriched skeletal stem cells. Biomaterials 2009;30(28):5094-5103. 166. Santos MI, Unger RE, Sousa RA, Reis RL, Kirkpatrick CJ. Crosstalk between osteoblasts and endothelial cells co-cultured on a polycaprolactone-starch scaffold and the in vitro development of vascularization. Biomaterials 2009;30(26):4407-4415. 167. Ripamonti U, Van Den Heever B, Sampath TK, Tucker MM, Rueger DC, Reddi AH. Complete regeneration of bone in the baboon by recombinant
205
human osteogenic protein-1 (hOP-1, bone morphogenetic protein-7). Growth factors (Chur, Switzerland) 1996;13(3-4):273-289,color plates III-VIII,pre bk. 168. Ramoshebi LN, Ripamonti U. Osteogenic protein-1, a bone morphogenetic protein, induces angiogenesis in the chick chorioallantoic membrane and synergizes with basic fibroblast growth factor and transforming growth factor-beta1. The Anatomical record 2000;259(1):97-107. 169. Schumann D, Ekaputra AK, Lam CX, Hutmacher DW. Biomaterials/scaffolds. Design of bioactive, multiphasic PCL/collagen type I and type II-PCL-TCP/collagen composite scaffolds for functional tissue engineering of osteochondral repair tissue by using electrospinning and FDM techniques. Methods in molecular medicine 2007;140:101-124. 170. Miura M, Miura Y, Padilla-Nash HM, Molinolo AA, Fu B, Patel V, et al. Accumulated chromosomal instability in murine bone marrow mesenchymal stem cells leads to malignant transformation. Stem cells (Dayton, Ohio) 2006;24(4):1095-1103. 171. Noth U, Rackwitz L, Heymer A, Weber M, Baumann B, Steinert A, et al. Chondrogenic differentiation of human mesenchymal stem cells in collagen type I hydrogels. J Biomed Mater Res A 2007;83(3):626-635. 172. Denker AE, Haas AR, Nicoll SB, Tuan RS. Chondrogenic differentiation of murine C3H10T1/2 multipotential mesenchymal cells: I. Stimulation by bone morphogenetic protein-2 in high-density micromass cultures. Differentiation; research in biological diversity 1999;64(2):67-76. 173. Haas AR, Tuan RS. Chondrogenic differentiation of murine C3H10T1/2 multipotential mesenchymal cells: II. Stimulation by bone morphogenetic protein-2 requires modulation of N-cadherin expression and function. Differentiation; research in biological diversity 1999;64(2):77-89. 174. Tortelli F, Tasso R, Loiacono F, Cancedda R. The development of tissue-engineered bone of different origin through endochondral and intramembranous ossification following the implantation of mesenchymal stem cells and osteoblasts in a murine model. Biomaterials;31(2):242-249. 175. Erlebacher A, Derynck R. Increased expression of TGF-beta 2 in osteoblasts results in an osteoporosis-like phenotype. The Journal of cell biology 1996;132(1-2):195-210. 176. Tatapudy S, Bruening S, Gleason G, Toth M. Validation and use of a computer-assisted counting procedure to quantify BrdU-labeled proliferating cells in the early postnatal mouse hippocampus. Journal of neuroscience methods 2008;172(2):173-177.
206
177. Roschger P, Fratzl P, Eschberger J, Klaushofer K. Validation of quantitative backscattered electron imaging for the measurement of mineral density distribution in human bone biopsies. Bone 1998;23(4):319-326. 178. Fratzl-Zelman N, Roschger P, Gourrier A, Weber M, Misof BM, Loveridge N, et al. Combination of nanoindentation and quantitative backscattered electron imaging revealed altered bone material properties associated with femoral neck fragility. Calcified tissue international 2009;85(4):335-343. 179. Chung ES, Saban DR, Chauhan SK, Dana R. Regulation of blood vessel versus lymphatic vessel growth in the cornea. Investigative ophthalmology & visual science 2009;50(4):1613-1618. 180. Wlodarski KH, Wlodarski PK, Galus R. Senescence of osteogenic cells. Review. Ortopedia, traumatologia, rehabilitacja 2007;9(1):63-67. 181. Hauschka PV, Lian JB, Cole DE, Gundberg CM. Osteocalcin and matrix Gla protein: vitamin K-dependent proteins in bone. Physiological reviews 1989;69(3):990-1047. 182. Hauschka PV, Wians FH, Jr. Osteocalcin-hydroxyapatite interaction in the extracellular organic matrix of bone. The Anatomical record 1989;224(2):180-188. 183. Yeh LC, Adamo ML, Olson MS, Lee JC. Osteogenic protein-1 and insulin-like growth factor I synergistically stimulate rat osteoblastic cell differentiation and proliferation. Endocrinology 1997;138(10):4181-4190. 184. Yeh LC, Tsai AD, Lee JC. Osteogenic protein-1 (OP-1, BMP-7) induces osteoblastic cell differentiation of the pluripotent mesenchymal cell line C2C12. Journal of cellular biochemistry 2002;87(3):292-304. 185. Pluhar GE, Turner AS, Pierce AR, Toth CA, Wheeler DL. A comparison of two biomaterial carriers for osteogenic protein-1 (BMP-7) in an ovine critical defect model. J Bone Joint Surg Br 2006;88(7):960-966. 186. Ristiniemi J, Flinkkila T, Hyvonen P, Lakovaara M, Pakarinen H, Jalovaara P. RhBMP-7 accelerates the healing in distal tibial fractures treated by external fixation. J Bone Joint Surg Br 2007;89(2):265-272. 187. Paul S, Lee JC, Yeh LC. A comparative study on BMP-induced osteoclastogenesis and osteoblastogenesis in primary cultures of adult rat bone marrow cells. Growth factors (Chur, Switzerland) 2009;27(2):121-131. 188. Cowan CM, Aghaloo T, Chou YF, Walder B, Zhang X, Soo C, et al. MicroCT evaluation of three-dimensional mineralization in response to BMP-2 doses in vitro and in critical sized rat calvarial defects. Tissue Eng 2007;13(3):501-512.
207
189. Liu SB, Hu PZ, Hou Y, Li P, Cao W, Tian Q. Recombinant human bone morphogenetic protein-2 promotes the proliferation of mesenchymal stem cells in vivo and in vitro. Chinese medical journal 2009;122(7):839-843. 190. Young S, Patel ZS, Kretlow JD, Murphy MB, Mountziaris PM, Baggett LS, et al. Dose effect of dual delivery of vascular endothelial growth factor and bone morphogenetic protein-2 on bone regeneration in a rat critical-size defect model. Tissue Eng Part A 2009;15(9):2347-2362. 191. Ljusberg J, Ek-Rylander B, Andersson G. Tartrate-resistant purple acid phosphatase is synthesized as a latent proenzyme and activated by cysteine proteinases. The Biochemical journal 1999;343 Pt 1:63-69. 192. Ljusberg J, Wang Y, Lang P, Norgard M, Dodds R, Hultenby K, et al. Proteolytic excision of a repressive loop domain in tartrate-resistant acid phosphatase by cathepsin K in osteoclasts. The Journal of biological chemistry 2005;280(31):28370-28381. 193. Minkin C. Bone acid phosphatase: tartrate-resistant acid phosphatase as a marker of osteoclast function. Calcified tissue international 1982;34(3):285-290. 194. Hayman AR, Jones SJ, Boyde A, Foster D, Colledge WH, Carlton MB, et al. Mice lacking tartrate-resistant acid phosphatase (Acp 5) have disrupted endochondral ossification and mild osteopetrosis. Development (Cambridge, England) 1996;122(10):3151-3162. 195. Angel NZ, Walsh N, Forwood MR, Ostrowski MC, Cassady AI, Hume DA. Transgenic mice overexpressing tartrate-resistant acid phosphatase exhibit an increased rate of bone turnover. J Bone Miner Res 2000;15(1):103-110. 196. Kaneko H, Arakawa T, Mano H, Kaneda T, Ogasawara A, Nakagawa M, et al. Direct stimulation of osteoclastic bone resorption by bone morphogenetic protein (BMP)-2 and expression of BMP receptors in mature osteoclasts. Bone 2000;27(4):479-486. 197. Hentunen TA, Lakkakorpi PT, Tuukkanen J, Lehenkari PP, Sampath TK, Vaananen HK. Effects of recombinant human osteogenic protein-1 on the differentiation of osteoclast-like cells and bone resorption. Biochemical and biophysical research communications 1995;209(2):433-443. 198. Deckers MM, van Bezooijen RL, van der Horst G, Hoogendam J, van Der Bent C, Papapoulos SE, et al. Bone morphogenetic proteins stimulate angiogenesis through osteoblast-derived vascular endothelial growth factor A. Endocrinology 2002;143(4):1545-1553. 199. Valdimarsdottir G, Goumans MJ, Rosendahl A, Brugman M, Itoh S, Lebrin F, et al. Stimulation of Id1 expression by bone morphogenetic protein
208
is sufficient and necessary for bone morphogenetic protein-induced activation of endothelial cells. Circulation 2002;106(17):2263-2270. 200. ten Dijke P, Fu J, Schaap P, Roelen BA. Signal transduction of bone morphogenetic proteins in osteoblast differentiation. J Bone Joint Surg Am 2003;85-A Suppl 3:34-38. 201. Chang SC, Tai CL, Chung HY, Lin TM, Jeng LB. Bone marrow mesenchymal stem cells form ectopic woven bone in vivo through endochondral bone formation. Artificial organs 2009;33(4):301-308. 202. Bastow ER, Byers S, Golub SB, Clarkin CE, Pitsillides AA, Fosang AJ. Hyaluronan synthesis and degradation in cartilage and bone. Cell Mol Life Sci 2008;65(3):395-413. 203. Hollister SJ. Scaffold engineering: a bridge to where? Biofabrication 2009;1(1):1-14. 204. Ratcliffe A. Industrial approaches in tissue engineering. In: Mao J, Vunjak-Novakovic G, Mikos AG, Atala A, editors. Translational Approaches in Tissue Engineering and Regenerative Medicine. Boston: Artech House Publishers, 2008. p. 463-472. 205. Konrad G, Sudkamp N. [Extra-articular proximal tibial fracture]. Der Chirurg; Zeitschrift fur alle Gebiete der operativen Medizen 2007;78(2):161-171; quiz 172-163. 206. Klein MP, Rahn BA, Frigg R, Kessler S, Perren SM. Reaming versus non-reaming in medullary nailing: interference with cortical circulation of the canine tibia. Arch Orthop Trauma Surg 1990;109(6):314-316. 207. Mueller CA, Rahn BA. Intramedullary pressure increase and increase in cortical temperature during reaming of the femoral medullary cavity: the effect of draining the medullary contents before reaming. The Journal of trauma 2003;55(3):495-503; discussion 503. 208. Wenda K, Ritter G, Degreif J, Rudigier J. [Pathogenesis of pulmonary complications following intramedullary nailing osteosyntheses]. Unfallchirurg 1988;91(9):432-435. 209. Pape HC, Hildebrand F, Krettek C, Green J, Giannoudis PV. Experimental background--review of animal studies. Injury 2006;37 Suppl 4:S25-38. 210. Baumgart F, Kohler G, Ochsner PE. The physics of heat generation during reaming of the medullary cavity. Injury 1998;29 Suppl 2:B11-25. 211. Giannoudis PV, Snowden S, Matthews SJ, Smye SW, Smith RM. Temperature rise during reamed tibial nailing. Clin Orthop Relat Res 2002(395):255-261.
209
212. Ochsner PE, Baumgart F, Kohler G. Heat-induced segmental necrosis after reaming of one humeral and two tibial fractures with a narrow medullary canal. Injury 1998;29 Suppl 2:B1-10. 213. Seekamp A, Lehmann U, Pizanis A, Pohlemann T. [New aspects for minimally invasive interventions in orthopedic trauma surgery]. Der Chirurg; Zeitschrift fur alle Gebiete der operativen Medizen 2003;74(4):301-309. 214. Ekeland A, Thoresen BO, Alho A, Stromsoe K, Folleras G, Haukebo A. Interlocking intramedullary nailing in the treatment of tibial fractures. A report of 45 cases. Clin Orthop Relat Res 1988(231):205-215. 215. Hansen M, Mehler D, Rommens PM. Das Verhalten winkelstabiler Implantatsysteme an der proximalen Tibia. Verlag Hans Huber, 2004. 216. Gautier E, Ganz R. [The biological plate osteosynthesis]. Zentralblatt fur Chirurgie 1994;119(8):564-572. 217. Krettek C, Schandelmaier P, Tscherne H. [New developments in stabilization of dia- and metaphyseal fractures of long tubular bones]. Der Orthopade 1997;26(5):408-421. 218. Gautier E, Perren SM, Cordey J. Strain distribution in plated and unplated sheep tibia an in vivo experiment. Injury 2000;31 Suppl 3:C37-44. 219. Hertel R, Eijer H, Meisser A, Hauke C, Perren SM. Biomechanical and biological considerations relating to the clinical use of the Point Contact-Fixator--evaluation of the device handling test in the treatment of diaphyseal fractures of the radius and/or ulna. Injury 2001;32 Suppl 2:B10-14. 220. Beardi J, Hessmann M, Hansen M, Rommens PM. Operative treatment of tibial shaft fractures: a comparison of different methods of primary stabilisation. Arch Orthop Trauma Surg 2008;128(7):709-715. 221. Weiss RJ, Montgomery SM, Ehlin A, Al Dabbagh Z, Stark A, Jansson KA. Decreasing incidence of tibial shaft fractures between 1998 and 2004: information based on 10,627 Swedish inpatients. Acta Orthop 2008;79(4):526-533. 222. Connolly JF, Guse R, Tiedeman J, Dehne R. Autologous marrow injection as a substitute for operative grafting of tibial nonunions. Clin Orthop Relat Res 1991(266):259-270. 223. Alberts KA, Loohagen G, Einarsdottir H. Open tibial fractures: faster union after unreamed nailing than external fixation. Injury 1999;30(8):519-523.
210
224. Darder-Garcia A, Darder-Prats A, Gomar-Sancho F. Nonreamed flexible locked intramedullary nailing in tibial open fractures. Clin Orthop Relat Res 1998(350):97-104. 225. Singer RW, Kellam JF. Open tibial diaphyseal fractures. Results of unreamed locked intramedullary nailing. Clin Orthop Relat Res 1995(315):114-118. 226. Whittle AP, Russell TA, Taylor JC, Lavelle DG. Treatment of open fractures of the tibial shaft with the use of interlocking nailing without reaming. J Bone Joint Surg Am 1992;74(8):1162-1171. 227. Swiontkowski MF, Senf D, Taylor S. Plate desgn has an effect on cortical bone perfusion. Trans 37th Orthop Res Soc, 1991. p. 387. 228. Korvick DL, Monville JD, Pijanowski GJ, Phillips JW. The effects of screw removal on bone strain in an idealized plated bone model. Vet Surg 1988;17(3):111-116. 229. Auer JA, Goodship A, Arnoczky S, Pearce S, Price J, Claes L, et al. Refining animal models in fracture research: seeking consensus in optimising both animal welfare and scientific validity for appropriate biomedical use. BMC Musculoskelet Disord 2007;8:72. 230. Kneser U, Polykandriotis E, Ohnolz J, Heidner K, Grabinger L, Euler S, et al. Engineering of vascularized transplantable bone tissues: induction of axial vascularization in an osteoconductive matrix using an arteriovenous loop. Tissue Eng 2006;12(7):1721-1731. 231. Mikos AG, Herring SW, Ochareon P, Elisseeff J, Lu HH, Kandel R, et al. Engineering complex tissues. Tissue Eng 2006;12(12):3307-3339. 232. Hing KA, Wilson LF, Buckland T. Comparative performance of three ceramic bone graft substitutes. Spine J 2007;7(4):475-490. 233. Schliephake H, Neukam FW, Hutmacher D, Wustenfeld H. Experimental transplantation of hydroxylapatite-bone composite grafts. J Oral Maxillofac Surg 1995;53(1):46-51; discussion 52. 234. Becker S, Maissen O, Ponomarev I, Stoll T, Rahn B, Wilke I. Osteopromotion by a beta-tricalcium phosphate/bone marrow hybrid implant for use in spine surgery. Spine 2006;31(1):11-17. 235. Leiggener CS, Curtis R, Muller AA, Pfluger D, Gogolewski S, Rahn BA. Influence of copolymer composition of polylactide implants on cranial bone regeneration. Biomaterials 2006;27(2):202-207. 236. Hutmacher DW. Scaffolds in tissue engineering bone and cartilage. Biomaterials 2000;21(24):2529-2543.
211
237. Hutmacher DW, Garcia AJ. Scaffold-based bone engineering by using genetically modified cells. Gene 2005;347(1):1-10. 238. Abbah SA, Lam CX, Hutmacher DW, Goh JC, Wong HK. Biological performance of a polycaprolactone-based scaffold used as fusion cage device in a large animal model of spinal reconstructive surgery. Biomaterials 2009;30(28):5086-5093. 239. Blair HC, Zaidi M, Schlesinger PH. Mechanisms balancing skeletal matrix synthesis and degradation. The Biochemical journal 2002;364(Pt 2):329-341. 240. Blumenthal NM, Koh-Kunst G, Alves ME, Miranda D, Sorensen RG, Wozney JM, et al. Effect of surgical implantation of recombinant human bone morphogenetic protein-2 in a bioabsorbable collagen sponge or calcium phosphate putty carrier in intrabony periodontal defects in the baboon. Journal of periodontology 2002;73(12):1494-1506. 241. Cheng H, Jiang W, Phillips FM, Haydon RC, Peng Y, Zhou L, et al. Osteogenic activity of the fourteen types of human bone morphogenetic proteins (BMPs). J Bone Joint Surg Am 2003;85-A(8):1544-1552. 242. Wu XB, Li Y, Schneider A, Yu W, Rajendren G, Iqbal J, et al. Impaired osteoblastic differentiation, reduced bone formation, and severe osteoporosis in noggin-overexpressing mice. The Journal of clinical investigation 2003;112(6):924-934. 243. Mikic B, van der Meulen MC, Kingsley DM, Carter DR. Long bone geometry and strength in adult BMP-5 deficient mice. Bone 1995;16(4):445-454. 244. Sykaras N, Opperman LA. Bone morphogenetic proteins (BMPs): how do they function and what can they offer the clinician? Journal of oral science 2003;45(2):57-73. 245. Janssens K, ten Dijke P, Janssens S, Van Hul W. Transforming growth factor-beta1 to the bone. Endocrine reviews 2005;26(6):743-774. 246. Linkhart TA, Mohan S, Baylink DJ. Growth factors for bone growth and repair: IGF, TGF beta and BMP. Bone 1996;19(1 Suppl):1S-12S. 247. Sawyer AA, Song SJ, Susanto E, Chuan P, Lam CX, Woodruff MA, et al. The stimulation of healing within a rat calvarial defect by mPCL-TCP/collagen scaffolds loaded with rhBMP-2. Biomaterials 2009;30(13):2479-2488. 248. Ugural AC, Fenster SK. Advanced Strength and Applied Elasticity. 3 ed. Englewood Cliffs, NJ: Prentice-Hall Inc., 1995.
212
249. Movat HZ. Demonstration of all connective tissue elements in a single section; pentachrome stains. A M A 1955;60(3):289-295. 250. Carter DR, Hayes WC. Bone compressive strength: the influence of density and strain rate. Science (New York, NY 1976;194(4270):1174-1176. 251. Yetkinler DN, McClellan RT, Reindel ES, Carter D, Poser RD. Biomechanical comparison of conventional open reduction and internal fixation versus calcium phosphate cement fixation of a central depressed tibial plateau fracture. J Orthop Trauma 2001;15(3):197-206. 252. Sanger C, Soto A, Mussa F, Sanzo M, Sardo L, Donati PA, et al. Maximizing results in craniofacial surgery with bioresorbable fixation devices. The Journal of craniofacial surgery 2007;18(4):926-930. 253. Hench LL, Polak JM. Third-generation biomedical materials. Science (New York, NY 2002;295(5557):1014-1017. 254. Hollister SJ. Porous scaffold design for tissue engineering. Nature materials 2005;4(7):518-524. 255. Lohmann H, Grass G, Rangger C, Mathiak G. Economic impact of cancellous bone grafting in trauma surgery. Arch Orthop Trauma Surg 2007;127(5):345-348. 256. Asahina I, Sampath TK, Nishimura I, Hauschka PV. Human osteogenic protein-1 induces both chondroblastic and osteoblastic differentiation of osteoprogenitor cells derived from newborn rat calvaria. The Journal of cell biology 1993;123(4):921-933. 257. Friedlaender GE, Perry CR, Cole JD, Cook SD, Cierny G, Muschler GF, et al. Osteogenic protein-1 (bone morphogenetic protein-7) in the treatment of tibial nonunions. J Bone Joint Surg Am 2001;83-A Suppl 1(Pt 2):S151-158. 258. Uludag H, Gao T, Porter TJ, Friess W, Wozney JM. Delivery systems for BMPs: factors contributing to protein retention at an application site. J Bone Joint Surg Am 2001;83-A Suppl 1(Pt 2):S128-135. 259. Erickson BP, Pierce AR, Simpson AK, Nash J, Grauer JN. 125I-labeled OP-1 is locally retained in a rabbit lumbar fusion model. Clin Orthop Relat Res 2008;466(1):210-215. 260. Uludag H, D'Augusta D, Golden J, Li J, Timony G, Riedel R, et al. Implantation of recombinant human bone morphogenetic proteins with biomaterial carriers: A correlation between protein pharmacokinetics and osteoinduction in the rat ectopic model. J Biomed Mater Res 2000;50(2):227-238.
213
261. Winn SR, Uludag H, Hollinger JO. Carrier systems for bone morphogenetic proteins. Clin Orthop Relat Res 1999(367 Suppl):S95-106. 262. Obert L, Deschaseaux F, Garbuio P. Critical analysis and efficacy of BMPs in long bones non-union. Injury 2005;36 Suppl 3:S38-42. 263. Takigami H, Kumagai K, Latson L, Togawa D, Bauer T, Powell K, et al. Bone formation following OP-1 implantation is improved by addition of autogenous bone marrow cells in a canine femur defect model. J Orthop Res 2007;25(10):1333-1342. 264. Bostrom M, Lane JM, Tomin E, Browne M, Berberian W, Turek T, et al. Use of bone morphogenetic protein-2 in the rabbit ulnar nonunion model. Clin Orthop Relat Res 1996(327):272-282. 265. Cook SD, Wolfe MW, Salkeld SL, Rueger DC. Effect of recombinant human osteogenic protein-1 on healing of segmental defects in non-human primates. J Bone Joint Surg Am 1995;77(5):734-750. 266. Geesink RG, Hoefnagels NH, Bulstra SK. Osteogenic activity of OP-1 bone morphogenetic protein (BMP-7) in a human fibular defect. J Bone Joint Surg Br 1999;81(4):710-718. 267. Haidar ZS, Hamdy RC, Tabrizian M. Delivery of recombinant bone morphogenetic proteins for bone regeneration and repair. Part A: Current challenges in BMP delivery. Biotechnology letters 2009;31(12):1817-1824. 268. Einhorn TA. Clinical applications of recombinant human BMPs: early experience and future development. J Bone Joint Surg Am 2003;85-A Suppl 3:82-88. 269. Dimitriou R, Tsiridis E, Giannoudis PV. Current concepts of molecular aspects of bone healing. Injury 2005;36(12):1392-1404. 270. Gerstenfeld LC, Alkhiary YM, Krall EA, Nicholls FH, Stapleton SN, Fitch JL, et al. Three-dimensional reconstruction of fracture callus morphogenesis. J Histochem Cytochem 2006;54(11):1215-1228. 271. Gerstenfeld LC, Cullinane DM, Barnes GL, Graves DT, Einhorn TA. Fracture healing as a post-natal developmental process: molecular, spatial, and temporal aspects of its regulation. Journal of cellular biochemistry 2003;88(5):873-884. 272. Le AX, Miclau T, Hu D, Helms JA. Molecular aspects of healing in stabilized and non-stabilized fractures. J Orthop Res 2001;19(1):78-84. 273. Cho TJ, Gerstenfeld LC, Einhorn TA. Differential temporal expression of members of the transforming growth factor beta superfamily during murine fracture healing. J Bone Miner Res 2002;17(3):513-520.
214
274. Yeh LC, Unda R, Lee JC. Osteogenic protein-1 differentially regulates the mRNA expression of bone morphogenetic proteins and their receptors in primary cultures of osteoblasts. Journal of cellular physiology 2000;185(1):87-97. 275. White AP, Weinstein MA, Patel T, Horowitz MC, Friedlaender GE. The 2002 Marshall Urist Young Investigator Award Paper. Lumbar arthrodesis gene expression: a comparison of autograft with osteogenic protein-1. Clin Orthop Relat Res 2004(429):330-337. 276. Onishi T, Ishidou Y, Nagamine T, Yone K, Imamura T, Kato M, et al. Distinct and overlapping patterns of localization of bone morphogenetic protein (BMP) family members and a BMP type II receptor during fracture healing in rats. Bone 1998;22(6):605-612. 277. Sampath TK, Maliakal JC, Hauschka PV, Jones WK, Sasak H, Tucker RF, et al. Recombinant human osteogenic protein-1 (hOP-1) induces new bone formation in vivo with a specific activity comparable with natural bovine osteogenic protein and stimulates osteoblast proliferation and differentiation in vitro. The Journal of biological chemistry 1992;267(28):20352-20362.
