On a biomechanical approach to analysis of stability and load bearing capacity of oral implants

(1)

MALMÖ UNIVERSITY isbn 978-91-7104-407-5 (print) isbn 978-91-7104-408-2 (pdf) ANDERS HALLDIN MALMÖ UNIVERSIT Y ON A BIOMEC HANIC AL APPR O A C H TO AN AL Y SIS OF S TABILIT Y AND LO AD BEARIN G C AP A CIT Y OF OR AL IMPL ANT S DOCT OR AL DISSERT A TION IN ODONT OL OG Y

ANDERS HALLDIN

ON A BIOMECHANICAL

APPROACH TO ANALYSIS

OF STABILITY AND LOAD

BEARING CAPACITY OF

ORAL IMPLANTS

(2)

(3)

O N A B I O M E C H A N I C A L A P P R O A C H T O A N A L Y S I S O F S T A B I L I T Y A N D L O A D B E A R I N G C A P A C I T Y O F O R A L I M P L A N T S

(4)

Malmö University

Faculty of Odontology Doctoral Dissertation 2015

(5)

ANDERS HALLDIN

ON A BIOMECHANICAL

APPROACH TO ANALYSIS

OF STABILITY AND LOAD

BEARING CAPACITY OF

ORAL IMPLANTS

Malmö University, 2015

Faculty of Odontology

Department of Prosthodontics

(6)

This thesis represents number 46 in a series of investigations on implants, hard tissue and the locomotor apparatus originating from the Department of Biomaterials, Universi-ty of Gothenburg and the Department of Prosthodontics/Material sciences, Malmö Uni-versity, Sweden.

1. Anders R Eriksson DDS, 1984. Heat-induced Bone Tissue Injury. An in vivo investiga-tion of heat tolerance of bone tissue and temperature rise in the drilling of cortical bone. Thesis defended 21.2.1984. External examiner: Docent K-G. Thorngren.

2. Magnus Jacobsson MD, 1985. On Bone Behaviour after Irradiation. Thesis defended 29.4.1985. External examiner: Docent A. Nathanson. 3. Fredric Buch MD, 1985. On Electrical Stimulation of Bone Tissue. Thesis defended 28.5.1985. External examiner: Docent T. Ejsing-Jörgensen.

4. Peter Kälebo MD, 1987. On Experimental Bone Regeneration in Titanium Implants. A quantitative microradiographic and histologic investigation using the Bone Harvest Chamber.

Thesis defended 1.10.1987. External examiner: Docent N. Egund.

5. Lars Carlsson MD, 1989. On the Development of a new Concept for Orthopaedic Implant Fixation.

Thesis defended 2.12.1989. External examiner: Docent L-Å Broström. 6. Tord Röstlund MD, 1990. On the Development of a New Arthroplasty. Thesis defended 19.1.1990. External examiner: Docent Å. Carlsson. 7. Carina Johansson Res Tech, 1991. On Tissue Reaction to Metal Implants. Thesis defended 12.4.1991. External examiner: Professor K. Nilner.

8. Lars Sennerby DDS, 1991. On the Bone Tissue Response to Titanium Implants. Thesis defended 24.9.1991. External examiner: Dr J.E. Davies.

9. Per Morberg MD, 1991. On Bone Tissue Reactions to Acrylic Cement. Thesis defended 19.12.1991. External examiner: Docent K. Obrant.

10. Ulla Myhr PT, 1994. On Factors of Importance for Sitting in Children with Cerebral Palsy.

Thesis defended 15.4.1994. External examiner: Docent K. Harms-Ringdahl. 11. Magnus Gottlander MD, 1994. On Hard Tissue Reactions to Hydroxyapatite-Coated Titanium Implants. Thesis defended 25.11.1994. External examiner: Docent P. Aspenberg.

12. Edward Ebramzadeh MScEng, 1995. On Factors Affecting Long-Term Outcome of Total Hip Replacements. Thesis defended 6.2.1995. External examiner: Docent L. Linder.

(7)

13. Patricia Campbell BA, 1995. On Aseptic Loosening in Total Hip Replacement: the Role of UHMWPE Wear Particles. Thesis defended 7.2.1995. External examiner: Pro-fessor D. Howie.

14. Ann Wennerberg DDS, 1996. On Surface Roughness and Implant Incorporation. Thesis defended 19.4.1996. External examiner: Professor P-O. Glantz.

15. Neil Meredith BDS MSc FDS RCSm 1997. On the Clinical Measurement of Implant Stability Osseointegration.

Thesis defended 3.6.1997. External examiner: Professor J. Brunski.

16. Lars Rasmusson DDS, 1998. On Implant Integration in Membrane-Induced and Grafter Bone.

Thesis defended 4.12.1998. External examiner: Professor R. Haanaes.

17. Thay Q Lee MSc, 1999. On the Biomechanics of the Patellfemoral Joint and Patel-lar Resurfacing in Total Knee Arthroplasty.

Thesis defended 19.4.1999. External examiner: Docent G. Nemeth.

18. Anna Karin Lundgren DDS, 1999. On Factors Influencing Guided Regeneration and Augmentation of Intramembraneous Bone.

Thesis defended 7.5.1999. External examiner: Professor B. Klinge.

19. Carl-Johan Ivanoff DDS, 1999. On Surgical and Impant Related Factors Influencing Integration and Function of Titanium Implants. Experimental and Clinical Aspects. Thesis defended 12.5.1999. External examiner: Professor B. Rosenquist.

20. Bertil Friberg DDS MDS, 1999. On Bone Quality and Implant Stability Measure-ments.

Thesis defended 12.11.1999. External examiner: Docent P. Åstrand.

21. Åse Allansdotter Johansson MD, 1999. On Implant Integration in Irradiated Bone. An Experimental Study of the Effects of Hyperbaric Oxygeneration and Delayed Implant Placement.

Thesis defended 8.12.1999. External examiner: Docent K. Arvidsson-Fyrberg. 22. Börje Svensson FFS, 2000. On Costochondral Grafts Replacing Mandibular Con-dyles in Juvenile Chronic Arthritis. A Clinical, Histologic and Experimental Study. Thesis defended 22.5.2000. External examiner: Professor Ch. Lindqvist.

23. Warren Macdonald BEng, MPhil, 2000. On Component Integration on Total Hip Arthroplasties: Pre-Clinical Evaluations.

Thesis defended 1.9.2000. External examiner: Dr A.J.C. Lee.

24. Magne Røkkum MD, 2001. On Late Complications with HA Coated Hip Asthro-plasties.

(8)

25. Carin Hallgren Höstner DDS, 2001. On the Bone Response to Different Implant Tex-tures. A 3D analysis of roughness, wavelength and surface pattern of experimental im-plants.

Thesis defended 19.11.2001. External examiner: Professor S. Lundgren.

26. Young-Taeg Sul DDS, 2002. On the Bone Response to Oxidised Titanium Implants: The role of microporous structure and chemical composition of the surface oxide in en-hanced osseointegration.

Thesis defended 7.6.2002. External examiner: Professor J.E. Ellingsen.

27. Victoria Franke Stenport DDS, 2002. On Growth Factors and Titanium Implant In-tegration in Bone.

Thesis defended 11.6.2002. External examiner: Associate Professor E. Solheim. 28. Mikael Sundfeldt MD, 2002. On the Aetiology of Aseptic Loosening in Joint Ar-throplasties and Routes to Improved cemented Fixation.

Thesis defended 14.6.2002. External examiner: Professor N. Dahlén.

29. Christer Slotte CCS, 2003. On Surgical Techniques to Increase Bone Density and Volume. Studies in the Rat and the Rabbit.

Thesis defended 13.6.2003. External examiner: Professor C.H.F. Hämmerle. 30. Anna Arvidsson MSc, 2003. On Surface Mediated Interactions Related to Chemo-mechanival Caries Removal. Effects on surrounding tissues and materials.

Thesis defended 28.11.2003. External examiner: Professor P. Tengvall.

31. Pia Bolind DDS, 2004. On 606 retrieved oral and cranio-facial implants. An analy-sis of consecutively received human specimens.

Thesis defended 17.12.2004. External examiner: Professor A. Piattelli.

32. Patricia Miranda Burgos DDS, 2006. On the influence of micro- and macroscopic surface modifications on bone integration of titanium implants.

Thesis defended 1.9.2006. External examiner: Professor A. Piattelli.

33. Jonas P Becktor DDS, 2006. On factors influencing the outcome of various tech-niques using, endosseous implants for reconstruction of the atrophic edentulous and par-tially dentate maxilla.

Thesis defended 17.11.2006. External examiner: Professor K.F. Moos. 34. Anna Göransson DDS, 2006. On Possibly Bioactive CP Titanium Surfaces. Thesis defended 8.12.2006. External examiner: Professor B. Melsen.

35. Andreas Thor DDS, 2006. On platelet-rich plasma in reconstructive dental implant surgery.

(9)

36. Luiz Meirelles DDS MSc, 2007. On Nano Size Structures For Enhanced Early Bone Formation.

Thesis defended 13.6.2007. External examiner: Professor Lyndon F. Cooper. 37. Pär-Olov Östman DDS, 2007. On various protocols for direct loading of implant-supported fixed prostheses.

Thesis defended 21.12.2007. External examiner: Professor B. Klinge.

38. Kerstin Fischer DDS, 2008. On immediate/early loading of implant supported pros-theses in the maxilla.

Thesis defended 8.2.2008. External examiner: Professor K. Arvidsson Fyrberg. 39. Alf Eliasson 2008. On the role of number of fixtures, surgical technique and timing of loading.

Thesis defended 23.5.2008. External examiner: Professor K. Arvidsson Fyrberg. 40. Victoria Fröjd DDS, 2010. On Ca2+ incorporation and nanoporosity of titanium surfaces and the effect on implant performance.

Thesis defended 26.11.2010. External examiner: Professor J.E. Ellingsen.

41. Lory Melin Svanborg DDS, 2011. On the importance of nanometer structures for implant incorporation in bone tissue.

Thesis defended 01.06.2011. External examiner: Associate professor C. Dahlin. 42. Byung-Soo Kang Msc, 2011. On the bone tissue response to surface chemistry modifications of titanium implants.

Thesis defended 30.09.2011. External examiner: Professor J. Pan.

43. Kostas Bougas DDS, 2012. On the influence of biochemical coating on implant bone incorporation.

Thesis defended 12.12.2012. External examiner: Professor T. Berglundh.

44. Arne Mordenfeld DDS, 2013. On tissue reactions to and resorption of bone substi-tutes.

Thesis defended 29.5.2013. External examiner: Professor C. Dahlin.

45. Ramesh Chowdhary DDS, 2014. On Efficacy of implant thread design for bone stimulation.

Thesis defended 21.05.2014. External Examiner: Professor Flemming Isidor 46. Anders Halldin Msc, 2015.On a biomechanical approach to analysis of stability and load bearing capacity of oral implants.

Thesis to be defended 28.05.2014. External Examiner: Professor J. Brunski. see www.mah.se/muep

(10)

(11)

ABSTRACT

Introduction

When an implant is placed in the bone the body responds to the trauma by encapsulating the implant and its survival depends on the ability for hard tissue encapsulation. The stability of the im-plant during the healing phase is essential to achieve a good result [1]. Biological, physiological and mechanical phenomena affect implant stability. To achieve sufficient stability during the initial healing phase the implant has to provide sufficient static interac-tion with the bone. The static interacinterac-tion might affect the biologi-cal processes that in turn affect implant stability. Although, nu-merous studies on the effect of dynamic interaction on implant sta-bility and bone remodeling exist, the effect of static strain has yet to be clarified.

As the healing progresses it may result in bone formation in close contact with the implant (i.e osseointegration) that stabilizes the implant. It has been found that implant surface modifications at the micro level promote osseointegration and that moderately roughened implants provide rapid and strong bone response [2, 3]. In addition, the application of nanostructures to an implant surface has been shown to elicit an initial complex gene response that may result in further enhancement in bone formation around the im-plant [4]. Furthermore the imim-plant surface structure interlocks me-chanically with the bone that affects the stability of the implant.

(14)

The implant surface design has to take into account both biological and mechanical behavior of the tissues.

Materials and methods

To investigate how implant stability and the biological response

are affected by an induced static load to the bone an in vivo study

was performed. Two types of controlled static loads, excessive and moderate, were induced by specially designed implants. Two types of surface structure, turned and blasted, were applied on the im-plants. The implants were inserted in rabbits and healed for 3-84 days before the stability was measured by removal torque.

To simulate how the pressure changes, due to biological and me-chanical phenomena, on an implant surface that was subjected to an initial pressure, a constitutive model was developed that was comprised of visco-elastic, visco-plastic and remodeling compo-nents. The pressure on the surface in turn affects the implant stabil-ity.

To investigate how the biomechanical and the biological responses

are affected by the surface structure an in vivo study and a finite

element analysis of the theoretical interfacial shear strength were performed. In the pre-clinical study, three groups of implants with different nano- and microstructures were compared to an implant with a control surface structure.

The theoretical interfacial strength at different healing times was estimated by simulating the surface structure interlocking capacity to bone using an explicit finite element method. Simulations were performed for different surface structures and for different pres-sures, simulating visco-elastic and remodeling phenomena.

(15)

Results

Implants that induced a moderate bone condensation in the bone had a significantly higher removal torque value at the implantation times of 3-24 days compared to implants that did not induce con-densation. The effect the induced moderate bone condensation had on implant stability decreases over time until the pressure has van-ished, which approximately occurred after 28-30 days. Turned im-plants, placed in tibia, that induced excessive bone condensation resulted in significant increased implant stability at implantation times of 3-24 days compared to implants that induced no conden-sation. However, when they were placed in femur it provided no significant difference in removal torque at an implantation time of 24 days compared to implants that induced no condensation. The developed constitutive model is able to capture visco-elastic material behavior and remodeling phenomena of cortical bone which can be used to simulate how the pressure changes on an im-plant surface that is subjected to an initial pressure caused by con-densation.

The implant nano- and microsurface structure affects the magni-tude of the removal torque value. It was found that implants, with

no significant difference in surface roughness parameters (Sa, Ssk,

Sdr) on micro level, can present a significant difference in removal

torque value at 4 weeks of implantation time. In addition, it was also found that implants with a significant difference in surface

roughness parameters (Sa, Ssk, Sdr) can present no significant

differ-ence in removal torque value at 4 weeks of implantation times. The difference may be due to various biological responses from the nano- and microstructure surfaces.

The simulated interfacial strength for the different surfaces did not reach the interfacial strength that corresponds to the removal

(16)

in respect of removal torque ratio, suggests that during the early healing phase the difference is caused by different bone formation rates from biological processes. As the healing progresses the effect of structural interlocking capacity is more pronounced.

Conclusions

The results suggest that increased static strain in the bone not only creates higher implant stability at the time of insertion, but also generates increased implant stability throughout the observation period of 3-24 days.

The proposed constitutive material model consists of three differ-ent compondiffer-ents: a visco-elastic compondiffer-ent, a visco-plastic compo-nent and a remodeling compocompo-nent. The model captures with good agreement the experimental behavior of cortical bone during

dif-ferent longitudinal loading situations i.e. in vitro stress-strain

rela-tionship, in vivo relaxation, in vitro creep and in vivo remodeling.

The results of the present study suggest that nano- and microstruc-ture alteration on a blasted implant might enhance the initial bio-mechanical performance, while for longer healing times, the sur-face interlocking capacity seems to be more important.

Simulation of the interfacial shear strength by means of finite ele-ment analysis seems to be a promising method to estimate the load bearing capacity of the bone-to-implant interface for different sur-face structures at stable healing conditions i.e. longer healing times. Furthermore, it is a promising method to estimate the implant sta-bility for different magnitudes of condensation.

(17)

INTRODUCTION

Bone is a complex material both from a biological, physiological and mechanical aspect. The biological structure and the physiolog-ical processes affect the geometrphysiolog-ical shape and mechanphysiolog-ical proper-ties of bone. The control mechanisms are complex but external forces have an important role in the control mechanisms that regu-late bone. One of the main functions of bone is to carry loads and work as stable skeletal support for the soft tissues. In addition it is also a protective material for sensitive organs and acts as a reser-voir for mineral for the organism. Over the last 50 years bone has been used to anchor implants of various functions. The success of implant function depends on stable anchorage in the bone. In the 1960s Per Ingvar Brånemark [5] showed that titanium implants can be used as anchorage devices due to their ability to maintain bone (hard tissue) in close contact to the implant under the right conditions which is termed osseointegration. Using the titanium implants as an anchoring device for a prosthetic structure was found to change the quality of life for people that were missing one or several teeth [6, 7].

(18)

LIST OF PAPERS

This thesis is based on 6 papers that will be referred to in the main text by their roman numerals. The papers are appended at the end of the thesis.

Papers

I. The effect of static bone strain on implant stability and

bone remodeling.

Anders Halldin, Ryo Jimbo, Carina Johansson, Ann Wen-nerberg, Magnus Jacobsson, Tomas Albrektsson and Stig Hansson. Bone 2011, 49(4), 783-789.

II. Implant stability and bone remodeling after 3 and 13 days

of implantation with an initial static strain

Anders Halldin, Ryo Jimbo, Carina Johansson, Ann Wen-nerberg, Magnus Jacobsson, Tomas Albrektsson and Stig Hansson.

Clinical Implant Dentistry and Related Research 2014 16(3): 383-393.

III. On a Constitutive Material Model to Capture Time

De-pendent Behavior of Cortical Bone.

Anders Halldin, Mats Ander, Magnus Jacobsson and Stig Hansson

(19)

IV. Improved osseointegration and interlocking capacity with dual acid-treated implants: a rabbit study

Anders Halldin, Ryo Jimbo, Carina Johansson, Christina Gretzer, Magnus Jacobsson

Clinical Oral Implants Research. doi: 10.1111/clr.12507

V. Simulation of the mechanical interlocking capacity of a

rough bone implant surface during healing.

Anders Halldin, Mats Ander, Magnus Jacobsson and Stig Hansson

Accepted in BioMedical Engineering OnLine. 16 April 2015

VI. Anders Halldin ,Yohei Jinno, Silvia Galli, Mats Ander, Stig

Hansson, Magnus Jacobsson, Ryo Jimbo

Implant stability and bone remodeling up to 84 days of

implantation with an initial static strain. An in vivo and

theoretical investigation.

Submitted to Journal of Clinical Periodontology. 2015 Reprint permission have been granted from

Paper I. Elsevier Limited License Number 3558280764833 Paper II. John Wiley and Sons License Number 3558281306677 Paper IV. John Wiley and Sons License Number 3558281107290 Paper III Open access

Contribution by the respondent

The respondent performed most of the work from planning,

exper-imental work, (except in vivo surgeries, surface characterizations,

cross sectioning and histomorphometric measurements), theoretical simulation and analysis of the data. The respondent was also the main contributor to writing the manuscripts.

(20)

PURPOSE OF THIS WORK

When an implant is placed in the bone the body responds to the trauma by encapsulating the implant and the implant survival de-pends on the ability for hard tissue encapsulation. The stability of the implant during the healing phase is essential to achieve a good result [1]. To achieve implant stability over time it is important to understand the biological, physiological and mechanical factors in order to fully utilize the capacity of bone tissue to incorporate for-eign bodies subjected to loads.

Bone is a biological material that undergoes changes to the materi-al characteristics over time due to both internmateri-al and externmateri-al bio-logical factors [8]. However, at a specific point in time it may be regarded as a material with unique structure and characteristics. It is therefore important to understand bone as a biological materi-al and how the hemateri-aling process affects its mechanicmateri-al properties at different points in time. Therefore we have conducted a series of tests and created a model in order to investigate the load bearing capacity of an implant over time.

x Studies I and II. In vivo measurements of the biomechanical

response over time to initial static strains. When an implant is placed in bone it may create, depending on implant design and corresponding drill progression, static bone strains that affect the biomechanical response and implant stability over time.

x Study III. Development of a constitutive material model to

(21)

surfaces in order to investigate the healing process and inter-locking capacity of structured surfaces.

x Study V. Development of a finite element model to estimate

the interlocking capacity of a rough implant surface with bone in order to simulate the implant stability during the osseointe-gration process.

x Study VI. Undertake in vivo measurements and simulation of

the biomechanical response and implant stability over time with an implant with a moderately rough surface that induces static bone strain at the time of insertion.

(22)

BONE AS BIOLOGICAL MATERIAL

Components of bone

Bone as a biological material consists mainly of collagen (20%) and mineral (65%) that surrounds the organic matrix [9]. The col-lagen, mainly of type I in bone, is a structural triple helix protein which can organize itself into strong fibers. The collagen gives bone its flexibility and tensile strength. At the end of the collagen molecules, bone mineral in form of hydroxy apatite Ca10(PO4)6(OH2) [10], provides the stiffness. A collagen fibril is built by a periodic arrangement of collagen molecules and hydroxy apatite. Several collagen fibrils build up the collagen fiber (figure 1) [11].

Figure 1. Collagen fiber. Modified after Rho et al. [11].

Subnano-structure Nanostructure Collagenfiber Collagenfibril Collagen molecule 1nm O.5μm

(23)

In addition, other substances exist in bone tissue such as proteogly-cans and non-collagenous proteins. It has been suggested that pro-teoglycans may modulate the collagen fibril assembly and control the organizing of the bone extracellular matrix [12]. The non-collagenous proteins, such as osteopontin, osteonectin, bone sialo-protein and osteocalcin, play an important role in initiation and control of the mineralization of bone [13]. Finally, bone also con-tains water, as well as supporting living cells and blood vessels.

Bone cells

Bone contains various types of specialized cells that are vital to maintain its biological balance. Numerous cells and chemical sig-nals interact in a complex process to maintain, repair or remove bone. Four main cell types can be recognized.

Osteoclasts

Osteoclasts are large multinucleated cells responsible for resorbing bone by creating a sealed attachment to the bone tissue to establish a microenvironment with increased acidity which enables local bone resorption [9, 10]. The cell surface inside the sealed attach-ment is extensively folded and is referred to as the ruffled boarder zone. Osteoclasts originate from stem cells found in the bone mar-row. These stem cells are differentiated through several steps into osteoclasts [14]. The final step in osteoclast formation involves fu-sion of mononuclear precursor cells into a multinuclear cell that is able to resorb bone [14]. In addition, fully developed monocytes and macrophages may also differentiate into osteoclasts [10]. The osteoclast stimulation and inhibition are regulated by numerous complex factors some of which are presented in table 1. osteopro-tegerin ligand (OPG-L) up-regulates differentiation, fusion and ac-tivation of osteoclast while osteoprotegerin (OPG) inhibits differ-entiation, fusion and activation of the osteoclast [10]. It has also been found that the Wnt signaling pathways (pathways made of proteins that pass signals from the outside to the inside of the cell) might suppress osteoclast activation through down-regulation

(24)

de-velopment of Receptor activator of nuclear factor kappa-B ligand (RANKL) [15]. Stimulation of osteoclasts to resorb bone may be initiated by inflammatory responses or several medical conditions such as cancer [16], osteoporosis [17] and bone damage [18]. Table 1. Stimulation and inhibitions factors of bone formation

Osteoblast

Osteoblasts are bone forming cells that synthesize and produce non-mineralized bone matrix (osteoid). The osteoid is comprised of approximately 90% collagen and 10 % non-collagenous proteins in which mineral will later be deposited [10, 19]. The osteoblast develops from mesenchymal stem cells (MC) found in marrow stroma [10]. Initially, the mesenchymal stem cells proliferate to be-come osteoblast precursor cells that in turn differentiate into pre-osteoblasts and subsequently mature pre-osteoblasts [15] (figure 2). The regulation of osteoblast differentiation is complex but is be-lieved to be caused by an essential transcription factor protein called core binding-factor-alpha (Cbfa1) or runt-related transcrip-tion factor 2 (RUNX-2) [10, 20] and by the Wnt signaling path-ways [15]. Active osteoblasts may follow one of three courses. They may become inactive and remain on the bone surface and be-come bone lining cells or they may surround themself with the bone matrix and become osteocytes or they may undergo apoptosis [9]. Furthermore, the osteoblast is essential for calcification and regulation of the calcium and phosphate balance in the organism.

(25)

Figure 2. The regulation of osteoclast and osteoblast differentia-tion. Modified after Baron et al. [21] and Goldring et al.[15] Bone lining cells

Bone lining cells are non-active osteoblasts that cover the surface of bone, by constructing a continuous sheet of cells that controls the ion exchange between the bone and the rest of the body [10, 19]. When the bone lining cells are present at the outer bone surface they are a part of the periosteum and when they are at the inside, i.e. lining the marrow cavity, they are a part of the endosteum. The bone lining cells are capable to form new bone without prior bone resorption, thus increasing bone mass [22]. The cells are a part of the mechanosensing mechanism that is influenced by functional strain to initiate new bone formation [23, 24]. In addition, bone lining cells are involved in the bone resorption process through di-gestion of the surface osteoid by proteases enabling osteoclast ac-cess to mineral [10]. Bone-lining cells represent a population of cells that is derived from inactive osteoblasts [9, 10].

Osteocytes

Osteocytes are abundant in mature bone and are derived from os-teoblasts incorporated in the bone during bone formation [9]. The

(26)

osteocytes form an extensive canalicular network connecting them to bone surface lining cells, osteoblasts, and other osteocytes. The canalicular network plays a role in the exchange of nutrients and waste in the bone. The network allows osteocytes, osteoblasts and bone lining cells to sense strain changes and microdamage to initi-ate, through chemical signaling, modeling and remodeling, [9, 24]. This mechanosensing ability is a prerequisite for the adaptation of the bone structure to external loads.

Woven and lamellar bone structure

Bone can be divided in into two types, namely woven bone and lamellar bone. Woven bone is comprised of a random matrix of collagen and mineral that develops rapidly (30-50 μm/day) [25]. Woven bone is found in the early stage of child skeletal growth (<3-5 years) [9, 26] and during early bone formation in fracture healing [27]. Woven bone is quickly mineralized in order to gain strength [26]. The mineralization process is random and simulta-neously impregnates the ground substance and collagen, resulting in a highly mineralized porous, bone structure [9, 19]. The random collagen-fibril orientation and irregular mineralization pattern of woven bone result in a more flexible, more easily deformed, and weaker bone than lamellar bone [9]. Woven bone is thereafter gradually remodeled to lamellar/Haversian bone by a remodeling process. The structure of Haversian bone is different from that of the woven bone. The Haversian bone is comprised of several indi-vidual lamellae, 3 to 7 μm thick, of collagen and mineral in defined directions, which are present in both cortical and trabecular bone. Lamellar bone is the principal load bearing tissue of the adult skel-eton and is formed slowly (0.6 μm/day) [25]. The lamellae are formed by a planar or cylindrical arrangement of mineralized col-lagen fibers. The colcol-lagen fibers lie in parallel in each lamella, but the orientation is different from one lamella to the next. This change of direction can be illustrated as a twisted "plywood" la-mellar structure of collagen fibrils [28].

(27)

Cortical bone structure

The cortical bone is a dense compact type of bone with a porosity of 5-10% and represents approximately 80% of the skeletal mass [10]. It is comprised of three different patterns of lamellar bone (figure 3) [10].

Figure 3. Structure of cortical bone modified from Cowin [26]. 1. Several circular rings of lamellae surrounding a central or

Ha-versian canal that builds up the individual osteon or HaHa-versian system.

2. Several layers of circumferential lamellae that build up the outer surface of the cortical bone.

3. Parallel interstitial lamellae that fill up the voids between oste-ons.

(28)

The osteon has a diameter of approximately 200 μm built up by approximately 20-30 circular lamellae [10]. The Haversian canals, approximately 50 μm in diameter, contain blood vessels, lymphatic vessels, and occasionally nerves. Central canals are connected by the transverse Volkmann’s canals. The central canals together with Volkmann’s canals contribute to the porosity of the cortical bone [9]. The cortical bone consists of several cylindrical oste-ons/Haversian systems, generally aligned with the principal loading direction [29]. The voids between the cylindrical osteons are filled with interstitial lamella.

Trabecular bone structure

The trabecular bone is different from cortical bone in the micro- and macrostructure. On a macro level trabecular bone is an inter-connected porous structure with a porosity of 50-90% [9] Trabec-ular bone is built up of several individual trabeculae characterized as rod or plate like structures containing several layers of lamellae. The thickness of the single lamella within the trabecular bone structure is in the range of 2-4 μm [30]. The thickness of the tra-beculum is approximately 50-100 μm and the length is

approxi-mately 1000 μm [19, 29]. It has been found that the trabecular

ori-entation is predominantly in the direction of the principal strains caused by the dominating loads [31].

(29)

Bone is a physiological material that changes its material properties over time due to biological activities. Some of the biological activi-ties are controlled or initiated by a complex interaction of genetics [32], chemical [10, 15, 24] and/or mechanosensing [18, 24, 31] processes. One objective of the biological activity is to optimize or maintain load bearing capacity by increasing or decreasing the bone mass, defined as bone modeling; restructure or repair bone damage on the microscale, defined as bone remodeling; and repair damaged bone on the macro scale (i.e. bone fractures), defined as bone fracture healing. The ability to sense mechanical stimuli is es-sential for modeling, remodeling and bone healing activities. It has been suggested that osteocytes act as a mechanotransducer of the mechanical forces on the tissue [33, 34]. This may occur through detecting local fluid flow within the bone [33, 34]. It has been found that osteocytes which are highly sensitive to the fluid shear forces within the canaliculi, secondary to changes in deformation of bone, could trigger chemical signals to the periosteal osteoblast layer [33, 35]. Frost [36, 37] developed a theory stating that mod-eling or remodmod-eling requires a certain level of deformation ex-pressed in strain to be initiated based on minimal effective strain (MES) (figure 4). MES below ~100 microstrain might lead to re-duction in bone mass (disuse atrophy), whereas steady state of

BONE AS A PHYSIOLOGICAL

MATERIAL

(30)

normal remodeling exists from ~100-~1000 microstrain and ~1000-~2000 microstrain represent net gain (physiological win-dow). Strains above ~3000 microstrain create microdamage and stress fracture. Peak load levels above 25000 microstrain cause fracture.

Figure 4. Minimal Effective Strain (MES). Modified from Frost [37].

Qin et al. [38] presented a similar theory according to which strains are necessary to maintain net bone mass based on daily stress or strain stimuli and the number of loading cycles per day (figure 5). According to Qin et al.’s [38] theory, decreased strain levels required a greater number of cycles to maintain bone mass. Strain levels above the threshold result in an increased bone mass

(31)

and strain levels below the threshold results in a decreased bone mass.

Figure 5. Microstrain threshold required to maintain bone mass as a function of daily loading cycle number [38].

It is well known that dynamic strains are the driving force behind bone modeling and remodeling [38, 39]. The knowledge about how static strain affects bone modeling and remodeling is limited

[40]. Few in vivo tests exist that can be used to extract the

remod-eling effect of static strain [41-44]. Perren et al. [41] induced a stat-ic pressure to intact and osteotomized sheep tibia bone and meas-ured the development of longitudinal forces over time. They con-cluded that after the initial relaxation of bone the pressure slowly

decreases linearly over time. The linear decrease in Perren’s [41] in

vivo experiment was explained by regular internal remodeling that

gradually removes pre-stressed bone and replaces it with non-pre-stressed bone. Similar findings on pre-non-pre-stressed bone were observed in two studies by Cordey et al. [43] and Blümline et al. [42]. In a study by Lynch and Silva, [45] it was concluded that bone damage in rat forelimb created by static creep loading can trigger modeling and woven bone response.

(32)

Modeling

Cortical and trabecular bone that are subjected to mechanical load have the ability, through mechanosensing, to adapt the structure to the mechanical loading by increasing or decreasing the bone mass. Modeling results in a change of bone mass either by bone apposi-tion or bone resorpapposi-tion. Bone apposiapposi-tion occurs when osteoblasts are recruited to the site to form osteoid which mineralizes over time to gain strength which results in new lamellar bone. The ap-position rates are in the range of 2-10 μm/day [10]. It has been found that a mechanical loading program enhanced the structural properties of a bone by bone adaptation at biomechanically rele-vant sites [46-48]. Bone resorption occurs when osteoclasts are re-cruited to the site. It has been found that disuse of bone might cause resorption [47] and that there is a threshold of minimum me-chanical stimuli required not to decrease bone mass [38, 49]. The rate of resorption is in the range of 0.15-7.3 μm/day [50-53].

Remodeling

Cortical and trabecular bone undergo continuous changes in struc-ture and composition, defined as remodeling (repair and reorienta-tion of structure) of bone. Remodeling is an important process to continuously repair, strengthen and adapt the bone to maintain load bearing capacity. Remodeling is also important to control the mineral balance in the body [25, 54]. This process is defined as a coupled process where resorption precedes apposition which sults in unchanged bone mass [25]. However, disruption in the re-modeling process might lead to an imbalance in bone mass [55]. In adult cortical bone the turnover rate is relatively low (2 to 10% yearly) [24, 54]. The turnover rate of trabecular bone is in the range of 20-30% which is more than is required to repair bone for maintaining bone strength, indicating that trabecular bone turno-ver is important for maintaining the mineral balance [24, 54]. In addition, the remodeling rate is species dependent where the rabbit remodeling rate is 3 times that of human [25].

(33)

Remodeling of cortical bone

Replacement of damaged and necrotic bone, by vital bone, is achieved by a transient functional group of osteoclasts and osteo-blasts entitled creeping substation and originally presented by Ax-hausen [56]. The process of creeping substitution was found to progress about 30 μm a day by Albrektsson [57] who used a direct vital microscopic technique and noticed that the so called cutting cones preferably invaded cortical bone by following old Haversian canals. In modern remodeling theories, creeping substation is re-ferred to as the basic multicellular unit (BMU) (figure 6).

Figure 6. Basic multicellular unit in cortical bone. Modified from Cowin [26].

The BMU is about 2 mm long and 0.2 mm wide and progresses at a rate of 20-40 μm/day longitudinally and 5-10 μm/day radially [10]. It preferentially moves in the direction of largest principal strain caused by the dominating loads [58]. The BMU cuts and re-builds its way through the bone to form a new Haversian system. The remodeling of cortical bone can be described in five phases: activation, resorption, reversal, formation and mineralization. Dur-ing the activation phase the osteoclasts need to first be recruited to the site [59]. The factors that initiate the activation are unknown but believed to be physical and biochemical signals which tie

oste-Matur e osto blast O_st oc las_t M atu_re ost_ob last Ost ocla st

(34)

oclast and osteoblast function together as a coupled event [55]. The osteoclast and the osteoblast may be looked upon as a balance with dominance of the one cell or the other, but with a strikingly different outcome; bone resorption or bone formation [60]. Apart from being bone cells, the osteoclast and the osteoblast are also a functioning part of the immune system [61].

In the resorption phase multicellular osteoclasts dissolve the bone at the front of the BMU, followed by the mononuclear osteoclast resorption to form a cavity. It has been suggested that the mono-nuclear cell resorption is responsible for digesting collagen [50]. The reversal phase is the time between resorption and bone for-mation and includes the recruitment of preosteoblasts which differ-entiate to osteoblasts. The reversal phase is in the range of 1-2 weeks [10].

In the formation phase the osteoblasts produce an osteoid which, after the mineralization lag time, is mineralized to gain bone strength [59, 62]. The mineralization lag time is in the range of 8-80 days for humans [63-66] and 1.6-3.7 days in rat and mouse [67-70]. Longer mineralization lag times result in increased accu-mulation of osteoid that leads to bone deficiency (osteomalacia) [8]. Fuchs et al. [71] studied the overall mineralization of the oste-oid in osteons over time. They found that the mineralization pro-cess can be divided into two stages: 1) primary mineralization and 2) secondary mineralization [71]. Primary mineralization is typified by a rapid linear rate of mineralization that proceeds until the re-modeling cavity has reached a degree of mineralization of 50-60% compared to interstitial bone (figure 7) [71, 72].

Following primary mineralization, the rate of mineralization slows and a phase of secondary mineralization progressively continues for a number of years in human. Mineralization is rarely, if ever, complete and typically stabilizes at around 90 to 95% of the max-imal theoretical level [72]. During this process the number of min-eral crystals increases and the size of the crystals is augmented [73].

(35)

Figure 7. Degree of mineralization of the osteoid in time measured in a single osteon [71].

(36)

Remodeling of trabecular bone

The process of remodeling of trabecular bone is similar to that of the cortical bone but with the difference that the trabecular bone remodels from the surface of the trabeculae and moving perpen-dicularly to the bone surface, culminating in a new bone structural unit (figure 8) [59]. The remodeling of trabecular bone can be de-scribed with the same 5 phases as cortical bone remodeling. In ad-dition, it has been found that the trabecular bone remodels to-wards the direction of main principal strain. [31].

Figure 8.Trabecular bone remodeling initiated by 1) activation 2a) resorption phase (a8 days) by osteoclasts followed by 2b) mononu-clear cell resorption (a34 days) 3) reversal or a preostoblastic phase (a34 days) 4) osteoid formation and mineralization lag time and 5) mineralization time. Modified from Cowin [26].

Bone healing

When bone fractures the biological processes start to repair and re-store the bone to its original function. The regeneration of tissue is an intricate function of the biological processes. It is a very com-plex process that involves the coordinated participation of migra-tion, differentiation and proliferation (figure 9) of inflammatory cells, angioblasts, fibroblasts, chondroblasts and osteoblasts which synthesize and release bioactive substances of extracellular matrix components (e.g., different types of collagen and growth factors) [74]. Mature osto blast Ostoclast Bonelining cell O sto_cl as_t Mo_non ucle ar cell_reso rptio_n 2a 2b 3 4 5 1

(37)

Figure 9. Sequential tissue differentiation processes of mesenchy-mal stem cells. Modified from Doblare et al. [74].

The healing process can be divided into primary or secondary frac-ture healing where the former is characterized by direct cortical re-construction and the latter characterized by a periosteal callus for-mation [75].

Primary healing process

Primary healing can occur when bone parts are in close contact and under rigorous stability which allow the BMUs to cross the site without woven bone formation [41, 76]. Consequently the fracture site heals by ordinary remodeling of the bone, resulting in longer healing time for the fracture.

(38)

Secondary healing process

Secondary healing occurs when there is insufficient stability and the gap size is moderate. Most bone fractures are repaired by sec-ondary healing, due to the nature of the fracture type and the com-plexity of providing a close rigid connection of the bone tissue part. Secondary fracture healing comprises a series of sequential stages that overlap to a certain extent, including inflammation, cal-lus differentiation, ossification and remodeling [74].

It is initiated by blood entering the site and quickly filling up the void (figure 10 a) followed by a sequential tissue differentiation processes to form an external callus (figure 10 b). The callus stabi-lizes the fracture by increasing the geometry and stiffness through formation of soft and hard callus tissue during the healing phase (figure 10 b, c). The increase in stability during healing decreases the movement in the fracture and woven bone formation may eventually occur in the gap (figure 10 d). Woven bone is thereafter mineralized to gain strength [10] resulting in further stabilization of the fracture. Finally, modeling and remodeling of the fracture site begins to restore the original structure and shape [74] (figure 10 d).

(39)

Figure 10. Secondary fracture healing has a series of sequential stages.

Mechanosensing of bone

It has been found that the loading and deformation conditions af-fect the type of tissue developed [75]. Several theories have been developed to explain the mechanisms that control bone morpho-genesis and tissue generation secondary to interfragmentary strain (figure 11 a-d). Pauwel’s theory for differentiation of mesenchymal cells into tissues, depending on the combination of volumetric and deviatoric deformation (strain) [77]. b) Perren and Cordey’s theory is based on critical strain limit for various tissue [78]. c) Carter et al. [79] proposed a model based on tensile strain and hydrostatic pressure. (d) Prendergast et al. [80] proposed tissue differentiation based on the relative fluid/solid velocity and tissue shear strain.

Bonemarrow Corticalbone Hematoma Disrupted periosterum Softtissuecallus, fibroustissue Bone formation Cartilage Remodelingof bonebyBMUs a b c d

(40)

Figure 11. Theories have been developed to explain the mecha-nisms that control bone morphogenesis and tissue generation. Fig-ures modified from a) [77] b) [78] c)[79] d)[80].

Healing process around a foreign body

During the surgical procedure of bone drilling and implant place-ment the host inevitably reacts to the trauma. The purpose of a foreign body reaction is to encapsulate the foreign body from the host. The healing process around a foreign body is complex, but includes blood–material interactions, provisional matrix formation, acute inflammation, chronic inflammation and tissue capsule de-velopment [81]. In the initial phase blood interacts with protein, adsorbed on the implant surface, to develop a blood-based provi-sional matrix around the implant leading to acute inflammation. Polymorphonuclear (PMNs) and mast cells affect the level of the acute inflammation. Following acute inflammation, chronic

in-Bone Connective tissue -0.15 Principaltensile strain% HydrostaticStress Fibrocartilage Cartilage < Bone

Bone2% Cartilage 10%Fibroustissue15% Granulationtissue 100%

Cart_ilag e Bone Reso_rption Time Tissueshe arstrain% Granulatio ntissue Fibroustis sue Deformation Compre ssion Endochondral ossification Hyalinca_rtilage Fib_ro usc art_ila ge Fib ro_u st issu_e Wovenbone Lamellarbone M_es en chy m al st_em ce_ll

a

b

c

d

Strain Rel at ive fluid/so lid ve lo ci ty Fibr_ous tiss_ue

(41)

flammation has been used to describe the reaction of a foreign body where monocytes, macrophages, and foreign body giant cells are present [81]. Adherent macrophages and foreign body giant cells in close proximity to the foreign body are known to increase the risk of implant failure [82]. The foreign body is separated from the encapsulation tissue by one or two layers of monocytes, mac-rophages, and foreign body giant cells [81]. The type of encapsula-tion tissue depends on biological and mechanical factors [75]. Al-brektsson et al. [82] have interpreted the foreign body reaction around an implant as a chronic inflammation. When the chronic inflammation progresses it may be appear as marginal bone resorp-tion and eventually be further compromised by plaque accumula-tion, infection and implant failure. However, encapsulation might be regarded as successful if the inflammatory response is limited and stable which is most often the case in clinical practice.

Bone healing around implants

A surgical trauma might result in traumatized (necrotic, damaged) bone [83] or voids in the bone. When a small gap or void exists af-ter a surgical trauma the healing process starts to fill the void with new tissue which is similar to the secondary bone healing process [84]. Bone formation during successful bone healing can briefly be summarized in two main processes: 1) formation of woven bone and 2) remodeling of woven bone to lamellar/Haversian bone. It has been found that a similar sequence of bone formation occurs in a defect size of 0.6, 1 and 10 mm [85, 86].

When an implant is inserted into bone the healing process starts by creating new bone that encapsulates the implant, called osseointe-gration. When a gap or void is present successful osseointegration and bone healing around an implant proceeds in the same way as secondary healing [84, 85, 87] with additional interfacial healing initiated from the surface by osteoblasts [88]. The interfacial heal-ing involves cellular and molecular responses that may be influ-enced by biomaterial surface texture, implant chemical

(42)

composi-tion, and implant biomechanics [88]. In a dog study Berglundh et al. [87], found woven bone after 1 week of implant healing which progressed and at 4 weeks woven bone could still be observed, of-ten together with parallel-fibered and lamellar bone. At 8-12 weeks signs of remodeling could be observed in the bone tissue. The re-moval of necrotic and damaged bone around an implant is per-formed by BMUs replacing the dead bone with vital bone [89]. Slaets et al. [89] found that the damaged bone had an increased in-vasion BMUs which reached its maximal after 4 weeks but was still active 6 weeks after implant insertion. In unsuccessful healing, the implant is encapsulated with soft connective tissue instead of bone tissue. Connective tissue proper might be a result of excessive preparation trauma, infection, or an overload that results in in-creased interfragmentary strain levels [78, 84].

To the authors’ knowledge the understanding of the healing pro-cess of an implant that is stable and in close contact with the bone is limited. Several studies have raised concerns that a press fitted implant may lead to bone necrosis and bone resorption [87, 90, 91], however no conclusive evidence has been presented and sever-al studies have shown no induced rapid bone remodeling due to induced press fit [43, 92-96]. Colnot et al. [88] studied the healing process around a 1 mm titanium cylinder inserted into a 1 mm hole to mimic close contact and press fit healing. The results showed that in some cases a small gap was present (<60 μm) and that the time-course of healing was equivalent whether or not a small gap existed. It was also found that the presence of an implant resulted in accelerated differentiation of peri-implant cells into osteoblasts, and acceleration in the remodeling of new bone matrix compared to an empty site [88].

(43)

BONE AS A PHYSICAL MATERIAL

The behavior of cortical bone subjected to load

The bone material can be described as a hierarchical system were all levels contribute to the mechanical properties of bone at the macro level [11] (figure 12).

Figure 12. Hierarchical structural of cortical and trabecular bone. Modified from Rho et al. [11].

Numerous investigations have been conducted to identify the de-formation behavior at the macro structural level of bone during loading in order to identify mechanical properties and develop ap-propriate constitutive material models [41, 97-106]. On the macro-structure level bone is able to adapt to different requirements with

Subnano-structure Nanostructure Submicro-structure Micro-structure Macro-structure Cancellousbone Corticalbone Osteon Trabeculae Lamella Collagenfiber Collagenfibril Collagen molecule 1nm O.5μm 3-7μm 10-500μm 50_-10 0μ_m ~10 00μ m 2-4 μm Haversiancanal

(44)

change in the structure and composition in turn affecting the me-chanical properties of the bone [19]. The measured meme-chanical properties, at the macro level, of bone are considerably affected by: 1) selection of specimen (species, location and orientation) 2) sam-ple condition (wet or dry) and 3) loading conditions (constant de-formation rate or loading rate) and test method (axial test, bending etc.) [19, 26]. Due to the variation of results this section will main-ly focus on the behavior of bone instead of describing obtained values found in the literature.

General stress-strain behavior

Mature bone has different mechanical properties in different direc-tions and is defined as an anisotropic material. However, in order to describe the mechanical properties of cortical bone, the material can be simplified and considered as a transverse isotropic or ortho-tropic material (figure 13) [11, 26, 105].

Figure 13. Longitudinal, transverse and radial direction of an or-thotropic bone material model.

A transverse isotropic material has different mechanical properties in two defined directions (i.e. composite) and an orthotropic mate-rial has different mechanical properties in three defined directions (i.e. wood). When a bone specimen is subjected to mechanical test-ing in the longitudinal direction the behavior can be illustrated in a stress-strain curve which represents a normalized load-deformation curve (figure 14). Under load bone initially exhibits a linear elastic

Tran_sve rse

(45)

behavior after which it reaches a yield point with an associated yield strain [97, 100]. Yield strains of cortical bone are approxi-mately 0.01 [107]. The linear relationship between stress and stain is defined by Young’s modulus. Young’s moduli are in the same range for both tension and compression but the yield strain is in-creased in compression compared to tension [26, 108]. Beyond the yield strain, post yield strain occurs in the bone. Post yield strain reflects damage to the material in the form of microcracks and/or internal slip between the mineral and collagen that degrade the ma-terial until failure [109, 110]. Beyond yield strain, bone may with-stand increased loading due to hardening before failure [97, 100, 111] until it reaches an ultimate strain of approximately 0.02 [107] before fracturing.

Figure 14. General behavior of bone loaded in longitudinal direc-tion. Obtained from McElhaney [97], Crowninshield and Pope [100] and Melnis and Knets [111].

When a bone specimen is subjected to mechanical loading in the transverse direction the behavior can also be illustrated in a stress-strain curve [100] (figure 15). In the transverse direction bone ex-hibits linear elastic property until it reaches a failure strain without

any yield strain [100].Reilly and Burstein [105] proposed a

(46)

tionship between longitudinal, transverse and radial Young’s mod-ulus of 1.6:1.3:1.0. while Qin et al. [38] obtained a ratio of 2.2:1.5:1.0 respectively.

Figure 15. General behavior of bone loaded in transverse direction obtained from Crowninshield and Pope [100].

It has been found that the stress-strain curve, in tension and com-pression, is affected by the magnitude of strain rate [97, 98, 100, 111-114]. In these studies the material stiffness increased with in-creased strain rate. The strain rate also affects the magnitude of the yield strain. The results provided by Melnis and Knets [111] indi-cated that a higher strain rate results in a higher yield strain. Simi-lar findings were obtained by Currey [115] and Crowninshield and Pope [100]. However, in a compression test McElhaney [97] found that a low strain rate did not affect yield strain level but that high strain rates result in decreased yield strain. Contradictory results were reported by Hansen et al. [113] where an increased strain rate resulted in decreased yield strain levels. These differences in results may reflect different testing procedures and environments such as, specimen sampling and specimen preparation. The strain rate also affects the evolution of post yield strain, where lower strain rate exhibits an increase of post yield strain before failure [97, 100,

(47)

109]. In the radial direction the Young’s modulus seems independ-ent of strain rate [100] and exhibits a brittle fracturing behavior without the development of post yield strain. An increased strain rate results in an increased fracture strain [100].

Relaxation and creep behavior

It has been found that bone exhibits relaxation [41, 116, 117] and creep behavior [118-120]. In a relaxation test the deformation (strain) is held constant and the force (stress) is measured over time. In a relaxation test the stress decreases with time. Iyo et al. [116] found that the relaxation phenomenon of a bovine femur bone consists of two different processes: a fast relaxation process

(relaxation time in the order of 102_{s) explained by relaxation of}

the collagen matrix and a slower relaxation process (relaxation

time in order of 106_{s) related to the higher order structure of the}

bone (figure 12). Similar slow relaxation responses might be

ob-served in the in vivo study on sheep by Perren et al. [41] (figure

16). In a creep test the force (stress) is held constant and the asso-ciated deformation (strain) is measured over time. In addition, if the force is removed the reverse creep deformation can be meas-ured over time. Depending on the initial force level a residual creep deformation might remain when the force is removed. In a creep experiment of human tibial bone, Melnis et al. [118] found that the bone exhibited both creep deformation, reverse creep deformation and residual creep. Similar findings were observed by Abdel-Wahab et al. [119] and Fondrk et al. [121].

(48)

Remodeling behavior

When living bone is under pre-stress, internal remodeling by the

BMU’s affects the stress level over time. In an in vivo experiment

Perren et al. [41] installed compression plates in sheep tibia and measured the development of longitudinal forces over time. They concluded that an initial decrease in axial force is related to the visco-elastic properties of bone and that a subsequent linear de-crease in pressure is related to remodeling (figure 16). The remod-eling process, which removes pre-stressed bone and substitute it with stress free bone, is a slow process which progresses over sev-eral weeks. Similar findings have been reported by Cordey et al. [43] and by Blümlein et al. [42].

Figure 16. Relaxation and remodeling behavior of cortical bone that affects the stress over time [41]. The deformation is held con-stant.

(49)

Variables affecting mechanical properties of bone

Bone is mainly built up of mineral collagen and water, which con-tribute to the mechanical properties of bone [9, 111, 120, 122, 123].

Mineral content

Bone constitutes a hierarchical structure, where all structural levels govern the macro material properties [11, 124]. Currey [123] found that Young’s modulus, in the direction of the long axis of the bone, is correlated to the amount of hydroxyapatite, where an increased content of the mineral results in a higher Young’s modu-lus. In addition, the same author found that bone seems to yield at a particular strain level of 0.0036-0.012 rather than at a particular stress level, with lower strain values for high mineral content and that the post yield strain decreases with increased values of mineral content. The degree of mineralization is affected by the type of bone (species, location) [123] which might be an explanation for the various results in Young’s modulus found in the literature. The behavior of bone is also affected by the properties (size, amount, crystallinity) of the mineral [72]. During secondary mineralization, there is a shift towards an increased number of crystals, as well as increased crystal size and crystallinity which results in increased brittleness failure [72].

(50)

Collagen and water

The amount of collagen and water affects the visco-elastic proper-ties of bone which results in different mechanical behavior for dif-ferent strain rates [111, 120, 125]. Bowman et al. [120] suggested that the collagen is responsible for the creep behavior of bone and Sasaki et al. [126] proposed that the collagen is responsible for the relaxation of bone. A similar suggestion was made by Iyo et al. [116]. In addition, it has also been proposed that the crosslinking of collagen affects the ultimate strain of bone [122].

Age related effects

It is known that the strength of bone deteriorates with age. Burstein et al. [127] found that mature bone decreases in strength and that stiffness decreases at a rate of ~2% per decade. It is there-fore important to understand how age-related changes in bone af-fect the material properties. It has also been found that age increas-es the porosity which in turn affects the macroscopic level mechan-ical properties, such as ultimate stress, ultimate strain, and yield stress [72, 128]. Hoffler et al. [129] studied the microscopic level mechanical properties of lamellar bone. Their results indicate that material properties are independent of age and gender. This sug-gests that the change in the mechanical properties may be an effect of changes on the macroscopic level, such as porosity etc. [72, 108]. In addition, increased age has demonstrated both increased crystal size and crystallinity, thus increasing the brittleness of bone [72].

(51)

Bone material mechanical model

When an implant is inserted into a smaller pre-drilled hole in the bone it creates strains with corresponding stresses in the bone that gradually decrease due to the mechanical and biological properties of bone (relaxation and remodeling) [41, 78]. In order to predict the stresses over time it is important to have a constitutive material model with parameters that capture the time dependent behavior of the material. Numerous material models have been developed where the individual models captures some of the behavior of bone subjected to a load [103, 130-133]. However, a material model will always be a simplification of the material it intends to repli-cate. In addition, the mechanical properties of bone are diverse [19, 26] consequently resulting in diverse model parameter values. However, some models that capture the behavior of bone need to be mentioned. Fondrk et al. [133] developed a visco-plastic damage model that could capture the nonlinear tensile behavior of cortical bone both in axial and bending loading. However, the biological reduction of pre-stress over time due to internal remodeling and the visco-elastic properties of bone were not included in the model. Johnson et al. [131] developed a visco-elastic and visco-plastic con-stitutive model that described the stress-strain relationship for dif-ferent strain rates. Their model demonstrates excellent agreement with published experimental tests results. However, neither the biological reduction of pre-stress over time due to internal remodeling nor relaxation nor creep nor hardening are included in the model developed by Johnson [131]. Garcia et al.[130] devel-oped a three-dimensional elastic plastic damage constitutive model which shows good agreement with their experimental test. This model seems to capture the degradation in stiffness due to repeated bone damage by cyclic behavior [108, 130]. However, this model does neither incorporate visco-elastic behavior nor remodeling.

(52)

OSSEOINTEGRATION OF DENTAL

IMPLANTS

The main function of a dental implant is to transfer the functional load from the tooth structure of the implant to the surrounding bone. The effectiveness of carrying loads depends on maintaining sufficient bone volume and bone strength. In addition, it has been shown that the maintenance of bone height is essential for tissue support for an enhanced esthetic outcome. In order to maintain sufficient bone support both biological and mechanical behavior of bone and how they interact with each other is essential. In addi-tion, the applied load should not negatively affect the ability to carry load during the patient’s lifetime. Dental implants transfer loads to the bone by the external threads added to the implant body. In later designs the surface roughness was modified to more effectively transfer shear forces [134, 135]. Once the implant is placed, it affects the biological environment and the biological tis-sue bone responses of the host. According to Albrektsson et al.[136] there are six factors that determine the tissue bone re-sponse to an implant placement:: 1) biocompatibility, 2) implant design, 3) implant surface, 4) state of the host bed, 5) surgical technique and 6 ) loading conditions. The ability to carry load over time thus depends on creating a biological response to main-tain sufficient bone support for the intended load.

(53)

Biocompatibility

In medical dictionaries [137-139] biocompatibility is commonly defined as the extent of not to cause a negative biological response, (such as inflammation, toxic, injurious, immune) to the biological system or the extent of a material to interact favorably with a bio-logic system. These classic definitions refer to the biobio-logical re-sponse caused by a foreign body. However, it does not incorporate the intended function of the foreign body. For a dental implant ap-plication it is essential that the implant is able to transfer physio-logical load to the bone. Williams [140] included the function in the definition: “Biocompatibility refers to the ability of a bio-material to perform its desired function with respect to a medical therapy, without eliciting any undesirable local or systemic effects in the recipient or beneficiary of that therapy, but generating the most appropriate beneficial cellular or tissue response in that spe-cific situation, and optimizing the clinically relevant performance of that therapy.” Using Williams’ definition, the ability of the im-plant to carry load over time is an important factor when evaluat-ing the biocompatibility of the implant. The ability to carry load for an implant depends on the strength of the material that is in contact with the implant over time. Several studies have shown that titanium has the material properties for creating and maintain-ing bone in close contact to the implant and to create sufficient bone support for the intended load capability. Other materials, such as zirconia or titanium alloys of various grades, are also used. Using Williams’ definition it is vital that the implant is designed to maintain bone in close contact to the implant over time to achieve load bearing capacity.

On a biomechanical approach to analysis of stability and load bearing capacity of oral implants

ANDERS HALLDIN

ON A BIOMECHANICAL

APPROACH TO ANALYSIS

OF STABILITY AND LOAD

BEARING CAPACITY OF

ORAL IMPLANTS

Malmö University

Faculty of Odontology Doctoral Dissertation 2015

ANDERS HALLDIN

ON A BIOMECHANICAL

APPROACH TO ANALYSIS

OF STABILITY AND LOAD

BEARING CAPACITY OF

ORAL IMPLANTS

Malmö University, 2015

Faculty of Odontology

Department of Prosthodontics

TableofContents

ABSTRACT

INTRODUCTION

LIST OF PAPERS

Papers

Contribution by the respondent

PURPOSE OF THIS WORK

BONE AS BIOLOGICAL MATERIAL

Components of bone

Bone cells

Woven and lamellar bone structure

Cortical bone structure

Trabecular bone structure

BONE AS A PHYSIOLOGICAL

MATERIAL

Modeling

Remodeling

Remodeling of cortical bone

Remodeling of trabecular bone

Bone healing

Primary healing process

Secondary healing process

Mechanosensing of bone

Healing process around a foreign body

a

b

c

d

Bone healing around implants

BONE AS A PHYSICAL MATERIAL

The behavior of cortical bone subjected to load

General stress-strain behavior

Relaxation and creep behavior

Remodeling behavior

Variables affecting mechanical properties of bone

Mineral content

Collagen and water

Age related effects

Bone material mechanical model

OSSEOINTEGRATION OF DENTAL

IMPLANTS

Biocompatibility