ACTAUNIVERSITATIS

UPSALIENSISUPPSALA

2018

Digital Comprehensive Summaries of Uppsala Dissertationsfrom the Faculty of Science and Technology 1676

Bone-compliant cements forvertebral augmentation

CÉLINE ROBO

ISSN 1651-6214ISBN 978-91-513-0351-2urn:nbn:se:uu:diva-349028

Dissertation presented at Uppsala University to be publicly examined in Häggsalen,Ångströmlaboratoriet, Lägerhyddsvägen 1, Uppsala, Thursday, 7 June 2018 at 09:00 forthe degree of Doctor of Philosophy. The examination will be conducted in English. Facultyexaminer: Professor Elizabeth Tanner (University of Glasgow).

AbstractRobo, C. 2018. Bone-compliant cements for vertebral augmentation. Digital ComprehensiveSummaries of Uppsala Dissertations from the Faculty of Science and Technology 1676. 92 pp.Uppsala: Acta Universitatis Upsaliensis. ISBN 978-91-513-0351-2.

Acrylic bone cement based on poly(methyl methacrylate) (PMMA) is commonly usedduring vertebral augmentation procedures for the treatment of osteoporosis-induced vertebralcompression fractures. However, the high stiffness of the cement compared to that of thesurrounding trabecular bone is presumed to facilitate the formation of new fractures shortlyafter surgery. The aim of the thesis was to develop and evaluate a PMMA-based bone cementthat better matches the mechanical properties of vertebral trabecular bone. To fulfill thisobjective, different compounds were added to the initial formulation of bone cement to modifyits functional properties. Linoleic acid (LA) was found to give the best combination of strengthand stiffness without negative effects on the handling properties and its use was therefore furtherinvestigated. In particular, different application-specific mechanical properties of LA-modifiedcement as well as itsin vivoperformance in an ovine model were assessed.

In summary, LA-modified cement exhibited bone-compliant mechanical propertiesimmediately after incorporation of the additive, as well as adequate handling properties, inparticular a lower polymerization temperature and appropriate setting time. The screw pulloutstrength from low-modulus cement was substantially reduced compared to regular PMMAcement, but comparable to some calcium phosphate based cements. The fatigue limit of LA-modified cement was considerably lower compared to regular PMMA bone cement when testedin physiological solution, but still higher than stresses measured in the spine during dailyactivities. The modified cement displayed similar inflammatory response in vivoto conventionalcement, with no evidence of additional cytotoxicity due to the presence of LA. Finally, it waspossible to sterilize the additive without significantly compromising its function in the PMMAcement.

The results from this thesis support further evaluation of the material towards the intendedclinical application.

Keywords: bone cement, low modulus, vertebroplasty

Céline Robo, Department of Engineering Sciences, Applied Materials Sciences, Box 534,Uppsala University, SE-75121 Uppsala, Sweden.

© Céline Robo 2018

ISSN 1651-6214ISBN 978-91-513-0351-2urn:nbn:se:uu:diva-349028 (http://urn.kb.se/resolve?urn=urn:nbn:se:uu:diva-349028)

To my parents, Alain and Patricia, and to Curiosity

There is no passion to be found playing small, in settling for a life that is less than the one you are capable of living.

- Nelson Mandela

List of Papers

This thesis is based on the following papers, which are referred to in the text by their Roman numerals.

I C. Persson, A. Robert, E. Carlsson, C. Robo, A. López, M. Go-

doy-Gallardo, M.-P. Ginebra, H. Engqvist. (2015) The effect of unsaturated fatty acid and triglyceride oil addition on the me-chanical and antibacterial properties of acrylic bone cements. Journal of Biomaterials Applications. 30(3): 279-289.

II C. Robo, D. Wenner, M. Nilsson, J. Hilborn, C. Öhman-Mägi, C. Persson. Functional properties of low-modulus PMMA bone cements containing linoleic acid. Manuscript.

III C. Robo, C. Öhman-Mägi, C. Persson. (2018) Compressive fa-

tigue properties of commercially available standard and low-modulus acrylic bone cements intended for vertebroplasty. Journal of the Mechanical Behavior of Biomedical Materials. 82: 70-76.

IV C. Robo, C. Öhman-Mägi, C. Persson. Long-term mechanical

properties of a low-modulus bone cement for the treatment of vertebral compression fractures. Manuscript.

V C. Robo, G. Hulsart-Billström, M. Nilsson, C. Persson. (2018)

In vivo response to a low-modulus PMMA bone cement in an ovine model. Acta Biomaterialia, e-pub ahead of print.

Reprints were made with permission from the respective publishers.

Author’s contributions

Paper I I contributed to some of the experimental work, particularly material preparation and mechanical characterization. I also contributed to the writing of the article.

Paper II I participated in the planning of the study and performed

most of the experimental work, except for the H-NMR and SFC-MS/MS analysis. I wrote the initial draft, contributed to the analysis of data, and was part of the continued writing process.

Paper III I participated in the planning of the study and performed all

the experimental work. I wrote the initial draft, contributed to the analysis of data, and was part of the continued writing process.

Paper IV I participated in the planning of the study and performed all

the experimental work, except for the HS-GC/MS analysis. I wrote the initial draft, contributed to the analysis of data, and was part of the continued writing process.

Paper V I participated in the planning of the study and performed all

the experimental work regarding material preparation and in vitro characterization, except for the HS-GC/MS analysis. I took part in the material preparation for the in vivo study. I wrote the initial draft, contributed to the analysis of data, and was part of the continued writing process.

Related publication by author

M. Pujari-Palmer, C. Robo, C. Persson, P. Procter, H. Engqvist. (2018) In-fluence of cement compressive strength and porosity on augmentation per-formance in a model of orthopaedic screw pull-out. Journal of the Mechani-cal Behavior of Biomedical Materials, 77:624-633.

Contents

1. Background ...........................................................................................131.1 The human spine .............................................................................14

1.1.1 Structure ..................................................................................141.1.2 Biomechanics of the spine and mechanical properties of vertebral bone ...................................................................................161.1.3 Osteoporosis and other spinal pathologies ................................181.1.4 Common treatments and treatment effects ................................20

1.2 Cements for spinal applications .......................................................231.2.1 History of acrylic bone cement .................................................231.2.2 Composition ............................................................................241.2.3 Polymerization Reaction ..........................................................251.2.4 Mechanical properties ..............................................................271.2.5 Low-modulus PMMA bone cements ........................................281.2.6 Other alternative ......................................................................30

2. Aims of the Thesis .................................................................................31

3. Materials & Methods .............................................................................333.1 Materials .........................................................................................33

3.1.1 Acrylic bone cements ...............................................................333.1.2 Additives .................................................................................333.1.3 Material preparation and synthesis ...........................................35

3.2 Methods ..........................................................................................363.2.1 Mechanical testing under quasi-static load ................................363.2.2 Fatigue properties.....................................................................383.2.3 Further characterization ............................................................403.2.4 Biocompatibility: in vivo evaluation .........................................42

4. Results...................................................................................................444.1 Handling properties .........................................................................444.2 DSC analysis ...................................................................................484.3 Quasi-static mechanical properties ...................................................484.4 Compressive fatigue properties ........................................................54

4.5 Additional material characterization ................................................564.5.1 1H-NMR and SFC-MS/MS analysis .........................................564.5.2 Monomer release ......................................................................574.5.3 Porosity ...................................................................................58

4.6 In vivo evaluation ............................................................................61

5. Discussion .............................................................................................64

6. Concluding remarks and Future perspectives..........................................73

Svensk Sammanfattning ............................................................................75

Acknowledgements ...................................................................................78

References .................................................................................................82

Abbreviations

σUCS

σy σb BaSO4 BPO DSC N.N-DMPT E Eb

FA Fmax

H-NMR LA Micro-CT MMA Nf

PBS PMMA SEM SFC-MS/MS Tg

Tmax

tsetting

VCF ZrO2

Ultimate compressive strength Yield strength Bending strength Barium sulfate Benzoyl peroxide Differential scanning calorimetry N.N-dimethyl-p-toluidine Elastic modulus Bending modulus Fatty acid Force at break Proton nuclear magnetic resonance spectrometry Linoleic acid Micro-computed tomography Methyl methacrylate Number of cycles to failure Phosphate buffered saline Poly (methyl methacrylate) Scanning Electron Microscopy Supercritical fluid chromatography-tandem mass Glass transition temperature Maximum polymerization temperature Setting time Vertebral compression fracture Zirconium dioxide

13

1. Background

Improved living conditions and better healthcare has resulted in a global population with an increased life expectancy. This, together with lower birth rates, has led to a higher proportion of elderly, which are also the ones typically affected by vertebral compression fractures (VCFs). These fractures commonly appear in patients suffering from pathologies such as osteoporosis, vertebral haemangioma, multiple myeloma or spinal metastasis. Hundreds of millions of people world-wide are affected by these types of fractures, incurring high societal costs and increased mortality rates within the population 1,2. Conservative techniques are conventionally suggested as primary means of treatment. However, the development of minimally invasive surgical procedures in the 1980s - vertebroplasty (VP) and balloon kyphoplasty (BKP) - has largely contributed to improving patients’ quality of life. These techniques consist in the percutaneous injection of a bone cement to provide pain relief and stabilization of the frac-ture. Although generally considered reliable and safe, these interven-tions can lead to complications related to the material injected into the vertebrae during surgery. This has given rise to an extensive amount of work dedicated to either finding new, alternative materials, or im-proving the current ones. The work presented in this thesis proposes a class of bone cements based on poly (methyl methacrylate) (PMMA) and focuses specifical-ly on the development of cements with bone-compliant mechanical properties. This is possible thanks to the addition of various com-pounds to the initial formulation of the cement. In this thesis the han-dling, quasi-static and dynamic mechanical properties, as well as the in vivo performance of the new formulations were investigated in or-der to further advance towards clinical use.

14

This thesis starts with a short section dedicated to the Human spine with a description of its Structure, Biomechanics and mechanical properties and related common Spinal pathologies such as Osteoporo-sis. Following this, the reader can find sections describing the differ-ent Cements developed for spinal applications and get an overview of the different alternatives. Next, the section on Materials and Methods describes the materials and analytical techniques used in this thesis. The section Results and Discussion presents the main findings of the work. The thesis ends with short a section on Concluding remarks and Future perspectives.

1.1 The human spine

1.1.1 Structure The human spine is made up of alternating vertebrae and interverte-bral discs held together by spinal ligaments and muscles (Figure 1) 3. It is composed of 24 articulated vertebrae and 9 fused segments form-ing the sacrum and the coccyx. The spinal column is divided into 5 regions (Figure 1 A):

1) The 7 vertebrae leading down from the top of the neck form the cervical spine, denoted C1-C7. These vertebrae support the weight of the head and allow neck movement (up and down as well as rotation to the left and right).

2) The 12 following vertebrae in the chest constitute the thoracic region, known as T1-T12. Their sizes vary depending on their location and they provide a limited range of motion. They sup-port the rib cage, which protects the lungs and heart.

3) The lower back of the spine is the lumbar region, composed of 5 segments denoted as L1-L5. They are the largest and strong-est vertebrae of the spine as they carry the weight of the upper body.

15

4) The sacrum, which consists of 5 segments during adolescence, forms a single block of bone in the adulthood. With the coc-cyx, it forms the terminal end of the spine attached to the pel-vis and connects the spine to the 2 hip bones.

5) The coccyx consists of the tailbone of the spine and is a single bone formed from the fusion of 4 tiny vertebrae. It serves as a focal point for the attachment of ligaments and muscles of the pelvic floor.

The whole spinal structure exhibits a natural S-shaped curve; concave (or lordosis) at the cervical and lumbar region, and convex (or kypho-sis) at the thoracic and sacral areas. These unique features of the spine (namely the combination of vertebrae and intervertebral discs of dif-ferent shapes and sizes) help to absorb shocks and ensure relative bal-ance of the spinal structure while in motion. A single vertebra is composed of two parts, the vertebral body (VB) and the vertebral (or neural) arch (Figure 1 B). The VB consists of a thin shell of compact bone (or cortical bone) surrounding a core of highly porous trabecular bone and red bone marrow (Figure 3). It is designed to withstand compressive forces (up to 80% of the body’s weight in an upright, standing position). The neural arches protect the spinal cord and provide attachment points for muscles and ligaments 4. The vertebrae are separated by intervertebral discs, which consist of an outer fibrocartilagenous ring known as annulus fibrosus, and a cen-tral gelatinous core known as nucleus pulposus 3. The main function of the discs is to facilitate the movement between vertebrae and dis-tribute the compressive forces evenly to the vertebral bodies 4. Vertebral endplates, found at the superior and inferior surfaces of the VB, are thin boundaries having an osseous as well as a cartilaginous component, that prevent extrusion of intervertebral discs into the po-rous trabecular bone of the VB and help to evenly transmit the loads to the vertebral body (Figure 1 B & C) 5.

16

Figure 1: Illustration of A) the 5 areas in the spinal column, B) an in-dividual vertebra from the top and C) from the side. (Reprinted from 3 with permission)

1.1.2 Biomechanics of the spine and mechanical properties of vertebral bone One of the main functions of bone cements is to reinforce the col-lapsed vertebra. Understanding the mechanical response of the verte-bral body during daily activities could therefore help to optimize the mechanical properties of the bone cement. Every physical activity (sitting, lifting a heavy object, walking, run-ning) creates loads on the spine. The vertebral bodies and interverte-bral discs support approximately 80% of the compressive forces expe-rienced by the spine, with the intervertebral ligaments resisting most of the bending forces 6. In a healthy individual, the compressive forces are distributed in the vertebral body between the trabecular bone - which bears the majority of the axial forces - and the shell of compact

Cervical

Thoracic

Lumbar

Sacral

Coccyx

C1

C7 T1

T12 L1

L5

A. B.

pedicle

C.

body

superior endplate

inferior endplate

spinous process vertebral foramen lamina

transverse process

discs

17

bone7. Any damage to the vertebral body or the intervertebral disc, or aged-related degenerative changes, markedly alter the load transfer as the loads are no longer evenly distributed in the vertebra. The load-bearing ability of the vertebral body depends, in addition to its size and shape, on the integrity of the trabecular system and bone mineral density (BMD)8. Therefore, changes in trabecular bone archi-tecture and decreases in BMD due to the normal aging process or bone pathology, compromises the strength of the trabecular bone and the structural capacity of the vertebral body 7. The strength and stiffness of vertebral trabecular bone have been measured in different studies and range between 0.1-15 MPa and 10-976 MPa respectively, with lower values for osteoporotic bone and higher for healthy bone 9,10. The stresses acting on the spine during normal daily activities, such as standing, walking or lying, do not ap-pear to exceed 3 MPa 11,12. Under repetitive cyclical loading due to normal daily physical activi-ties, the VB is subject to cumulative fatigue microdamage, resulting in a gradual loss of stiffness and strength of the VB 13. Different ex vivo studies have investigated the fatigue behavior of bovine and human vertebral trabecular bone using general linear regression or finite ele-ment models 14–16. However, comparison with the in vivo conditions is difficult as different factors may influence the fatigue behavior of bone. For instance, in vitro experiments preclude biological healing due to bone remodelling. Indeed, it has been hypothesized that fatigue damage might stimulate bone remodelling in vivo 17. Besides, the fa-tigue strength of bone depends on the age of the individual, frequency and strain rate, anatomic site, density and microstructure. The number of cycles to failure was found to be related to the applied stress divid-ed by the initial elastic modulus of the tested specimen.15 However, this relationship could not fully explain the fatigue behaviour. It has been demonstrated that fatigue failure of trabecular bone occurs at the ultrastructural level and the importance of creep in the mechanisms of deformation of trabecular bone during cyclic loading have been re-ported to be similar to cortical bone 14,15. Despite these findings, the fatigue behaviour of trabecular bone is still not fully understood.

18

1.1.3 Osteoporosis and other spinal pathologies Osteoporosis is the most common skeletal disease affecting primarily postmenopausal women and aging individuals of both sexes. This bone pathology is characterized by low bone mass and micro-architectural deterioration of the bone tissue, which in turn leads to bone fragility and an increased risk of fracture 18. This disease is gen-erally clinically silent and osteoporotic patients commonly do not dis-play any symptoms or signs of bone loss until painful osteoporotic fractures occur. These are most likely to occur in the spine, hip, distal forearm and proximal humerus. The most common kind of fracture is the vertebral compression fracture (VCF) 18, as illustrated in Figure 2. VCFs are defined as a reduction in height of an individual vertebra by at least 15% 19. The incidence of these fractures increases with age in both genders above the age of 50, although the risk is higher in wom-en than in men (approx. 1 in 2 compared to 1 in 5) 1. Spinal fractures have direct consequences on a patient’s quality of life as they cause spinal deformity (kyphosis) and severe back pain due to spinal misa-lignment, spasm of paraspinal muscles and ligament stretching 18. In addition, they are associated with an increased risk of mortality 20. In the USA alone, osteoporosis affects approx. 10 million people over the age of 50 and causes 1.5 million osteoporotic fractures each year, of which 50% are VCFs 18,21. In Europe, 27.5 million people over 50 are reported to have osteoporosis. Among this group, approximately 520 000 fractures (out of a total 3.5 million osteoporotic fractures) have been identified as located in the spine 22. Osteoporotic fracture treatment has a high societal cost, which is likely to increase with the growing number of ageing individuals worldwide 23. Other less common bone pathologies may also be associated with ver-tebral compression fractures. Multiple myeloma is a malignant hema-tologic tumor, which results in the formation of lytic bone lesions 24,25. In 2014, about 24 000 cases of multiple myeloma were diagnosed in the USA 24,25.

19

Intraosseous haemangioma is a benign tumor that develops from blood vessels and can affect the spine, in particular the thoracic and lumbar region of the vertebral body 26,27. The incidence of vertebral haemangioma is about 10-12% in adults 27. The spread of a cancer (such as breast, prostate, kidney and lung) from its site of origin to the vertebral bone is known as spinal metastasis 28. Spinal metastasis often results in bone loss due to lytic lesions, some-times with subsequent neurological deficit and severe back pain due to vertebral fractures 29. It affects people from all ages (mostly between 40 and 65) and develops mainly in the thoracic (60-80%) and lumbar spine (15-30%) 29. Data on the mechanical properties of pathological vertebral bone are scarce and often linear regression is used to correlate the bone fraction volume or bone mineral density of healthy cancellous bone to get an estimation of their strength and stiffness 10,30. Nazarian et al.10 esti-mated the stiffness (E) and the yield strength (σy) of both metastatic and osteoporotic trabecular bone (mixed vertebral and femoral trabec-ular bone). They found that osteoporotic trabecular bone had a Young’s modulus and yield strength comprised between 2-270 MPa and 0.1-7 MPa, respectively. The mechanical properties of metastatic bone were estimated to be slightly higher (E=40-640 MPa and σy = 0.1-24 MPa) 10. In short, these studies demonstrated that the mechani-cal properties of pathological bone can be considerably reduced com-pared to healthy cancellous bone.

20

Figure 2: Illustration of a vertebral compression fracture related to osteoporosis

1.1.4 Common treatments and treatment effects Medical therapies of VCFs were mostly nonsurgical before the devel-opment of vertebral augmentation techniques. Conservative care com-bines bed rest, pain control medication (narcotic analgesics, muscle relaxants and non-steroidal anti-inflammatory drugs), physiotherapy (with exercises) and back bracing for several weeks 21,31. These treat-ments are primarily aimed at pain relief and improving bone and mus-cle quality to reduce the risk of future fractures. A gradual reduction of pain is generally observed in the patients after a few weeks (2 -12 weeks) 32. However, these methods are not without complications or side effects. Prolonged immobilization of the patient can result in ac-celerated bone loss and muscle weakening. Additionally, the intensive use of medications can result in gastrointestinal irritation, renal insuf-ficiency and an increased risk of cardiac events in patients with hyper-tension and coronary artery disease 21,31,33. There are no standards specifying the length of conservative care. However, vertebral augmentation techniques are recommended when

Vertebral compression fractures

21

the previous methods are inefficient at relieving unremitting pain re-lated to VCFs (by approx. 3-6 weeks) and when patients continue to demonstrate fracture progression 31. Vertebroplasty (VP) and balloon Kyphoplasty (BKP) are two mini-mally invasive percutaneous procedures prescribed for the treatment of severe VCFs (minimum 15-20% height loss). These techniques help to consolidate the fracture and restore lost vertebral height in or-der to provide pain relief and stability. Percutaneous transpedicular vertebroplasty is a fluoroscopically guid-ed surgical intervention during which a radiopaque bone cement (typi-cally PMMA bone cement) is injected into a collapsed vertebral body. The cement is usually injected percutaneously via a cannula (10-15 G), using a bipedicular approach (through each pedicle) to provide an even distribution of the cement 34. This method was first reported by Deramond et al. 35 in 1987 to treat aggressive vertebral haemangioma and has become a standard of care in treating VCFs. Percutaneous balloon Kyphoplasty was later introduced by Garfin, Reilley and Lieberman et al. 36,37 for treating kyphotic deformities. It is similar to VP, except that it involves the insertion of an inflatable balloon tamp into the fracture prior to cement injection. This addition-al step helps to create a cavity for the bone cement. The inflation of the balloon tamp - whose pressure should not exceed 300 psi 38 - al-lows restoration of the vertebral body height and correction of the ky-phosis by realigning the spine36, although the long-term effect has been discussed39,40. From a clinician’s perspective, BKP also has the advantage to reduce the risks of cement extravasation in the vicinity (commonly into the epidural space, disk space or vascular system). These treatments have not been without controversy. In 2009, two randomized controlled trials 39,40 were published stating that, in terms of pain relief and quality of life, VP performed no better than a place-bo. These reports were met with a high level of criticism, and shortly afterwards, conflicting reports arose within the literature. With the exception of these two trials, the majority of studies support the effi-cacy of VP compared to nonsurgical management41.

22

The success of VP and BKP these last decades can be explained by the immediate pain relief (within 24h and up to 80-90% of the cases) experienced by the patients following surgery42–45. Even though the exact mechanism responsible for the palliative effects is unclear, it is believed that mechanical stabilization of the vertebral body gives the main effect, but also chemical and thermal necrosis of the neural tis-sues could contribute to pain relief 45. However, these interventions are also associated with some potential complications. First, cement leakage after vertebral augmentation (≤ 1 ml 46), although rare and usually asymptomatic, may cause local dete-rioration of tissue (chemical toxicity, thermal damage, spinal cord compression) or result in serious complications, in particular embolic phenomena 47–49. In addition, new fractures may form after the treat-ment. While these may occur as a result of the underlying disease, a disproportionally high number occur in the vertebrae adjacent to the treated ones 50–54. It is therefore believed that the high stiffness of the cement - which may act as a stress-raiser - and/or high injection vol-umes may cause, or at least facilitate, the occurrence of new fractures. Common indications for vertebral augmentation include painful oste-oporotic VCFs refractory to conservative treatments, painful vertebrae due to extensive osteolysis and benign bone tumors such as aggressive haemangomia or multiple myeloma38,55,56. Common absolute contra-indications concern patients with reporting improvement in symptoms after conservative care or with asymptomatic VCFs. In addition, it is not recommended to perform VP and BKP in patients with uncorrect-able coagulopathy, vertebra plana (completely collapsed vertebral bodies), severe cardiorespiratory disease and showing allergy to bone cement or contrast agent38,55,56.

23

Figure 3: Illustration of a bipedicular injection of bone cement per-formed during A) vertebroplasty and B) balloon kyphoplasty surger-ies. The patient (beige line) is placed in supine position and the ce-ment is injected within the vertebra, in the trabecular bone containing bone marrow.

1.2 Cements for spinal applications

1.2.1 History of acrylic bone cement Poly(methyl methacrylate) (PMMA) was first developed by the chem-ist Dr. Otto Röhm in 1901, within the scope of his thesis entitled “Polymerization products of acrylic acid” 57. This material was further

A.

B.

bone cement

Lumbar supine

vertebra

inflatable balloon tamp

trabecular bone

cannula

cortical shell

24

developed in his company “Röhm and Hass” and was marketed as PLEXIGLAS in 1933. They produced PMMA-based products for den-tures as well as submarine periscopes, windshields, canopies and gun turrets in airplanes during World War II. In 1936, Kulzer and Degussa discovered that the dough formed by mixing ground PMMA powder and a liquid monomer hardens when benzoyl peroxide is added and heated to 100°C 57. A few years later, they showed that this reaction could also take place at room temperature when a co-initiator (activa-tor) is added 57. This study established the founding principles for the production of PMMA bone cement. The first in vivo use of PMMA dates back to 1938, to an attempt to close cranial defects in monkeys. Kleinschmitt later repeated this pro-cedure in 1941 to close cranial defect in humans. Since then, the use of PMMA as bone cement has been extended to other fields including dentistry, ophthalmology and orthopaedics. In the 1940s, Jean and Robert Judet 58 introduced a method for fixating prosthetic implants using the cement, followed in the 1950s by Kiaer and Haboush 59, who used it in hip replacements. However, its current success and populari-ty for use in hip replacements is commonly accredited to Sir John Charnley 60. In the mid-1980s, Deramond et al. 35 started to use it for vertebral augmentation procedures (vertebroplasty and balloon ky-phoplasty), treating hemangioma-induced vertebral compression frac-tures (VCFs). Despite the extensive use of PMMA cement in vertebral augmentation, the Food and Drug Administration (FDA) did not ap-prove this material until 2004 61. The success of PMMA bone cement can be attributed to its bio-inertness, relatively good biocompatibility over long-term follow-up, and cost-effectiveness. Furthermore, it is easy to handle and provides good mechanical strength.

1.2.2 Composition PMMA bone cement is formed from 2 separate components, a liquid and a solid phase (Table 1). Even though the exact formulation differs

25

between brands, the cement usually consists of similar basic compo-nents. The solid phase, in the form of a powder, is generally composed of pre-polymerized MMA or methacrylate copolymer beads, an initia-tor for radical polymerization (typically, a peroxide) and a high amount of radiopacifier or contrast agent (typically, an inorganic ox-ide, 30-45 wt%). The liquid phase comprises mainly methyl methacry-late monomers (MMA) (sometimes butyl methacrylate) and the acti-vator (typically, a tertiary amine) to form radicals. Additionally, hy-droquinone is added to prevent premature polymerization during stor-age. Other additives consist of colorants (e.g. chlorophyll), used to distinguish cement from bone, and/or hydroxyapatite to enhancebio-compatibility of the cement or antibiotics. The amount of liquid and solid varies depending on the brand and is defined as the liquid-to-powder ratio (L/P ratio). Table 1: Common composition of PMMA bone cement

Powder Liquid Polymer PMMA, MMA/C Monomer MMA Initiator BPO Activator N.N-DmpT

Radiopacifier BaSO4, ZrO2 Inhibitor hydroquinone

Antibiotic Gentamicin Vancomycin

Additive chlorophyll, plasti-cizer

Additive coloring agent, plasticizer

1.2.3 Polymerization Reaction Bone cement is synthesized via a free-radical mechanism that occurs when the powder and liquid are mixed together. The initiator, benzoyl peroxide, is decomposed at room temperature by the activator N.N-DmpT via a reduction/oxidation reaction to form reactive benzoyl

26

radicals. These radicals initiate the polymerization by reacting with the C=C double bonds of the MMA monomer, as seen in Figure 4. The polymeric chains are then generated, growing as the number of radicals increases, and resulting in the formation of a dough with in-creased viscosity (molecular weight varying from 100 000 to 800 000 Da 57). The viscosity and reaction rates increase quickly due to the presence of PMMA particles. Chain mobility becomes limited as the cement hardens, so complete conversion of all MMA is therefore not possible. This generally leaves 2-6% of residual monomer in the ce-ment matrix. The dough typically sets in 15 min. During the process, heat is generated, resulting in temperatures of up to 120 °C.

Figure 4: Polymerization reaction

!"

!"!"

!" !"#$

%$

#!"$

#!"$

BPO N.N-DmpT

+

!"

!"

Radical (R )

"

!"

Radical (R )

1. Initiation (radical formation) 2. Chain growth

3. Chain recombination

R R +

COOCH3

C

CH3

H2C

MMA COOCH3

C

CH3

H2C R C

COOCH3

C

CH3

H2C + n

COOCH3

C

CH3

R

H3COOC

CH3

H2C

n

Radical

CH2 C

CH3

C )

COOH3

R )( H2C n 2

CH3

C

H3COOC

R ( H2C n C

COOCH3

CH3

CH2 ) R

27

1.2.4 Mechanical properties The mechanical properties of PMMA bone cement have been widely investigated 57,62,63. They are highly dependent on the components that comprise the cement formulation as well as the L/P ratio, but can also be affected by parameters such as temperature, porosity, storage con-ditions and the amount of additives incorporated into the cement for-mulation. Most of the mechanical tests used to characterize PMMA bone cement are performed according to the ISO 5833 standard 64. Some of the main quasi-static mechanical properties of PMMA ce-ments are summarized in Table 2. PMMA bone cements are known to be weaker in tension (≈ 35 MPa) than in compression (≈ 93 MPa) (Table 2) 65. For spinal applications, the most relevant properties are those measured in compression since the vertebral body is mainly sub-jected to quasi-static and repeated compressive loads during daily ac-tivities, such as standing, sitting or walking 6,11. The effect of repeated loading cycles on a material, i.e. fatigue, has been measured for PMMA cements in tension, compression and bending in previous work 66–68. Previously, there was no standard available for the evalua-tion of the fatigue properties of vertebroplastic cement. More recently, a modified version of the ASTM standard developed for acrylic bone cements used in fixating artificial joints (ASTM 2118-09) has been specified 69. It is indicated therein that the fatigue life of orthopaedic bone cements for spinal applications should be determined up to 5 million cycles at stress levels of 5, 7, 9 MPa in uniaxial tension-compression applied loading. In the current work, the cements were submitted to compression-compression loading since the cement will mainly experience compressive forces in the vertebroplastic applica-tion in vivo 6. Studies on the fatigue properties of standard PMMA bone cements intended for VP are scarce. The few studies available have reported fatigue limits varying between 6.8-10.4 MPa when the cement was submitted to fully reversed tension-compression 67,70,71 and 55 MPa when tested in compression-compression 66.

28

Table 2: Mechanical properties of PMMA bone cements compared with the properties required in ISO 5833 standard and those of verte-bral trabecular bone 57,62,72,73

Mechanical Properties [MPa]

PMMA

bone cement ISO standard (min value)

Vertebral trabec-ular bone

Compressive strength 70 – 110 70 1 – 15 Compressive modulus 1700 – 3000 - 10 – 976

Bending strength 50 – 80 50 - Bending modulus 1500 – 3000 1800 - Tensile strength 27 – 50 - 1.3 – 3.5 Tensile modulus 1580 – 4100 - 140 – 620

1.2.5 Low-modulus PMMA bone cements Low-modulus PMMA cements are currently being developed in an attempt to prevent the formation of new fractures reported to occur shortly after vertebral augmentation 50,53,74. The mechanical properties of commercially available PMMA bone cement can be reduced by incorporating additives 75,76 within the formulation to create mi-cropores 77,78 in the polymer matrix (dissolution), or to interfere with the polymerization reaction (plasticizers) 79. A plethora of formula-tions have been suggested, but to the best of the author’s knowledge, only one such cement is commercially available (Resilience®, Teknimed). This cement relies on the dissolution of an additional phase in the material, creatingporosity and lower mechanical proper-ties over time in vivo. This means that a low-modulus is not achieved until after few weeks (Paper III). Table 3 summarizes some of the oth-er approaches taken, together with reported mechanical properties (all aiming at a stiffness in the range of vertebral trabecular bone, E = 10-976 MPa). It has been reported that low-modulus cement may be ben-eficial for the restoration of the initial properties of fractured vertebral bodies 80–82. However, none of these experimental cements have reached clinical use.

29

Table 3: Experimental low-modulus acrylic bone cements intended for vertebral augmentation procedures.

Base cement Additives (amounts) Compressive mechanical properties (ranges)

Other mechan-ical properties

Vertecem™ (Synthes)78 sHA (aqueous frac-tion of the cement: 15-50 %)

E = 47- 929 MPa

σy = 1.3 - 39 MPa

Non-medical grade PMMA (Beracyl, Troller)83

HPMC (aqueous fraction of the ce-ment: 40-50 %)

E = 120 - 600 MPa

σUCS = 8 - 30 MPa

PEMA-nBMA84,85

E = 580 - 830 MPa Eb = 835 MPa

σUCS = 33.5 - 46.1 MPa σb = 29.3 MPa

Et = 650 MPa

σt = 19.3 MPa

Osteopal®V (Heraeus)86 LA (1.5 wt%)

E = 872 MPa

σUCS = 17.1 MPa

σy = 13.2 MPa

Vertecem (Synthes)87 NMP (aqueous frac-tion of the cement: 60 %)

E = 76-320 MPa σy = 2-24 MPa

Osteopal®V (Heraeus)88 CO (12 wt%) E = 446 MPa

σy = 15 MPa

Osteopal®V (Heraeus)89 oTMC (18-20 wt%) E = 241 - 734 MPa

σy = 5.2 - 25.6 MPa

Vertecem V+ (Synthes)90 SBF (8mL added to the cement) E = 955 MPa

Experimental formulation79 LA (15 v/v) E = 774 MPa

σUCS = 49 MPa

sHA = Sodium hyaluronate; HPMC = Hydroxypropylmethyl cellulose; PEMA-nBMA = Polyethylmethacrylate n-butylmethacrylate; LA = Linoleic acid NMP = 1-methyl-2-Pyrrolidone; CO = Castor oil; oTMC = oligo(trimethylenecarbonate); SBF: Fetal bovine serum; E = elastic modulus; Eb = bending modulus, Et = Tensile modulus; σy = Yield strength; σUCS = Ultimate compressive strength; σb = Bending strength; σt = Tensile strength.

30

1.2.6 Other alternative Examples of other potential alternatives to PMMA cement in verte-broplasty include composite and calcium phosphate cements (CPCs). Composite cements such as Orthocomp or Cortoss (Orthovita) are bioactive glass-ceramic cements with good biocompatibility and low setting temperature. However, these cements have similar mechanical properties to regular PMMA cements 91,92. Some commercial CPCs such as BoneSource (hydroxyapatite cement, Stryker) have been evaluated for treating fractured vertebra 93. Unlike PMMA cement, CPCs can resorb over time to be replaced by new bone. They are highly biocompatible, osteoconductive and set at low temperatures, preventing any risk of thermal necrosis of bone tissue. Other commercially available CPCs that have been evaluated for po-tential use in VP and BKP include Norian SRS (Synthes), Biopex (Mitsubishi) and KyphOs FS (Kyphon) 94,95. However, to the best of the author’s knowledge, only one ceramic cement is currently ap-proved for use in the spine, namely CERAMENT™ SPINE SUP-PORT. However, since these cements degrade over time in the body to be replaced by the patient’s own bone, real success of the treatment is dependent on the clinicians being able to also cure the underlying disease in the case of osteoporotic patients for example, and this is not evident. Furthermore, these types of cements have a poor load-bearing capacity compared to PMMA cements, in particular under other load-ing scenarios than purely compressive95.

31

2. Aims of the Thesis

Acrylic bone cement based on poly(methyl methacrylate) (PMMA) has been used in vertebroplasty since the 1980s and has become the material of choice in the field of spine augmentation due to its ease of handling and long-term mechanical performance. However, several drawbacks have been described, and in particular the increased risk of additional vertebral compression fractures (VCFs) post-surgery, which has been associated to the high stiffness of the cement. A PMMA bone cement with more bone-like mechanical properties is therefore a promising alternative in view of a more reliable cement performance, and the main aim of this thesis was to develop and evaluate the func-tional properties of such a cement. Consequently, a first aim of this thesis was to investigate the functional properties of new formulations of PMMA bone cements in terms of quasi-static mechanical analysis and handling properties. Besides immediate reinforcement of the ver-tebral body, the new material, in direct contact with the bone tissue and cells, must provide an adequate in vivo response. It will also be subjected to repetitive cyclic loads in vivo and must therefore be able to provide long-term support, as well as withstand the in vivo envi-ronmental conditions. There are many publications on regular PMMA cements; however, little is known about the in vivo properties and long-term performance of bone-compliant PMMA bone cements. As such, two other aims of this work were to evaluate the biocompatibil-ity and long-term mechanical properties of low-modulus PMMA bone cement. The present work focused on the following specific aims:

• To develop and evaluate new formulations of low-modulus PMMA bone cements containing various ad-

32

ditives in terms of mechanical and antibacterial proper-ties (Paper I)

• To evaluate the flexural properties, screw holding pow-er as well as the effect of additive sterilization on the properties of experimental low-modulus PMMA bone cements (Paper II)

• To compare the fatigue properties of commercially available regular and low-modulus PMMA bone ce-ments (Paper III)

• To evaluate the long-term mechanical performance of an experimental low-modulus PMMA bone cement in a physiological environment (Paper IV)

• To evaluate the in vivo performance of an experimental low-modulus PMMA bone cement in an ovine model (Paper V)

33

3. Materials & Methods

3.1 Materials

3.1.1 Acrylic bone cements Acrylic bone cement intended for clinical use is supplied in a sterile package containing the powder and the liquid phase (in a vial). The powder contains pre-polymerized polymer beads, a radiopacifier and an initiator. The liquid phase consists mainly of the methyl methacry-late monomer and the activator. In this thesis, five commercially available PMMA bone cements were used and a detailed description of the composition of these is given Table 4.

3.1.2 Additives The different additives utilised to modify the mechanical properties of commercially available PMMA bone cements are presented in Tables 5 and 6. These additives are unsaturated fatty acid compounds (FAs) (Papers I, II, IV, V and VI) and fatty-acid derived triglyceride oils (TOs) (Paper I). Although all the additives modified the mechanical properties of acrylic bone cement (Paper I), this thesis focused primarily on the unsaturated FA linoleic acid (LA), since it appeared to have a greater effect on the final mechanical properties of the cement compared to TOs, even when low amounts of LA were used 96.

34

Table 4: Composition of the commercially available PMMA bone cement used in this thesis (information provided by the supplier). Components are expressed in w/w except where stated and the liquid to powder ration (L/P ratio) is in ml/g.

Formulation

V-Steady™ Resilience® Osteopal®V F20®

Powder PMMA 54.15 15.2 - 19.2

P(MMA/MA) - 54.6 -

MMA/SC - 28.1 - 35.3

BPO 0.8 0.3 0.38 0.5

BaSO4 - -

ZrO2 45 39.2 45 45

HA - 4.3 - -

PAA - 12.9 - -

Colourant (E141) - - NS -

Liquid MMA 98.5 99 97.9 99.3

N.N-DMPT 0.014 1 2 0.7

HQ 50 ppm 20 ppm NS 20 ppm

Colourant (E141) - - NS -

L/P ratio 0.385 0.32 0.385 0.387

Viscosity High Medium Low Medium

Supplier G21 Teknimed Heraeus Teknimed

Paper II, IV III, IV I III, V

PMMA = Poly(methyl methacrylate); P(MMA/MA) = Poly(methyl methacrylate, methyl acrylate); MMA/SC = methyl methacrylate/styrene copolymer; BPO = ben-zoyl peroxide; BaSO4 = Barium sulphate; ZrO2 = Zirconium dioxide; HA = Hydrox-yapatite; PAA = poly(amino acid); MMA = Methyl methacrylate; N.N-DMPT = N,N dimethyl-p-toluidine; HQ = Hydroquinone; NS = Not specified.

35

Table 5: The fatty acids used in this thesis 97,98.

Name Formula Number of double bonds Paper

Linoleic acid C18H32O2 2 I, II, IV, V, VI

Ricinoleic acid C18H34O3 1 I

Methyl Linoleate C19H34O2 2 I

Table 6: Fatty acid contents (wt%) of the different oils used in Paper I 98–100.

Fatty acid Castor oil Linseed oil Tung oil Palmitic acid 0.8 -1.1 5.0 - 7.0 < 3.0 Stearic acid 0.7 - 1.0 3.0 - 6.0 < 4.0 Oleic acid 2.2 - 3.3 14.0 - 24.0 3.5 - 12.7

Linoleic acid 4.1 - 4.7 14.0 - 19.0 8.0 - 10.0 Linolenic acid 0.5 - 0.7 48.0 - 60.0 < 3.0 Ricinoleic acid 87.7 - 90.4 - -

α-Elaecostearic acid - - 77.0 - 84.0

3.1.3 Material preparation and synthesis The powder and monomer were mixed manually at room temperature, in a glass mortar for 30-45 seconds as recommended by the manufac-turers. Modified bone cements (Papers I, III, IV, V) were prepared by first mixing the additive with the monomer phase for a few seconds in a vortex to allow the formation of a homogenous liquid phase. It should be noted however, that this process is not needed for the linole-ic acid, which rapidly dissolves in the monomer liquid upon addition. Directly after monomer/additive mixing, the modified liquid phase was added to the powder and mixed as previously described. The pol-ymeric paste was then transferred to the moulds for each specific me-

36

chanical test or characterization technique and the cement was set in phosphate buffered saline solution at 37°C.

3.2 Methods

3.2.1 Mechanical testing under quasi-static load The acrylic bone cements in this thesis are designed for spinal applica-tions and may therefore be subject to compressive forces during daily physical activities 6. Two different mechanical tests were used to evaluate the quasi-static mechanical properties of the cements. Sche-matic drawings of the experimental configurations are shown in Fig-ure 5. The most common mechanical testing was compression (Paper I, II, III, IV and V). Cylindrical specimens (Ø = 6 mm; h = 12 mm) were compressed in the axial direction until failure occurred (by pass-ing the upper yield point) as recommended in ISO 5833 101. The ulti-mate compressive strength (σUCS= Fmax/A; Fmax=maximum applied load before specimen failure, A=original cross-sectional area) and elastic modulus (Young modulus, E, slope at 0.2-0.4% strain) were determined from the stress-strain curves as shown in Figure 6. In paper II, a four-point bending test was carried out as recommended in ISO 5833 101. Rectangular specimens (length = 75 mm; width = 10 mm; depth = 3.3 mm) were bent until failure and the deflection be-tween 15N and 50N was recorded and used to calculate the bending modulus (Eb). The bending strength (σb) was calculated using the force at break of the samples. E, σUCS, Eb and σb are expressed in megapascals (MPa).

37

Figure 5: Illustration of the experimental configuration for the quasi-static mechanical tests used in this thesis. The specimens were sub-jected to (1) compression and (2) bending loads. Tests were performed according to ISO 5333101.

Figure 6: Typical Stress-Strain curve of PMMA bone cement obtained after loading in compression. is the yield compressive strength and

is the ultimate compressive strength.

(1) (2)

0

140

0 0.

Stre

ss (M

Pa)

Strain (%)

E modulus = Δσ/Δε

ε

σ

σy σUCS

38

3.2.2 Fatigue properties Once implanted into the spine, the bone cement is expected to remain permanently in the vertebral body. A correct transfer of loads to the bone is therefore very important. The in vivo environment consists in static and dynamic stresses on the biomaterial, which will affect its performance over time. It is therefore important to understand and predict the material’s long-term behaviour under dynamic loading. The changes in properties and structural damage occurring in bone cement due to repeated cyclical loading were investigated by uniaxial fatigue testing. In papers III and IV, fatigue analysis was carried out based on ASTM F2118-10 69. In both studies, the specimens (Ø = 6 mm; h = 12 mm) were submitted to dynamic compressive fatigue loading until run out taken at 2 (Paper IV) and 5 (Paper III) million cycles or when the specimens failed, defined as a loss of 15% of the original height. A frequency of 2 Hz was used and the cements were either tested in air at room temperature (23 ± 0.1°C) or in saline solution at 37°C, the latter to better simulate the in vivo conditions. Specimen selection was based on macropore sizes, as recommended by ASTM F2118-10 69, where specimens containing internal defects > 1 mm were rejected. Since the first work on fatigue properties of acrylic bone cement by Krause and Mathis (1987), various methodologies have been suggest-ed over the years to perform fatigue tests and analyse the data 102. The most common types of applied loading are uniaxial tension-tension 103,104, uniaxial fully reversed tension-compression 67,103,105,106, uniaxial 3-point or 4-point bending 68,107 and cyclic compression-compression 66,108. The fatigue data are often presented as a plot of the Wölher curve (applied stress S versus number of cycles to failure Nf, or log Nf) 68,103,109 or analysed using fitting to 2- and 3-parameter Weibull rela-tions 67,103,106,110. In addition to these models, the Olgive-type equation used in Paper III and IV, has been widely used to mathematically rep-resent the S-Nf data 66,70,105,110,111. The Olgive-type equation may be expressed as follows:

39

A, B, C and D are the material constants and can be estimated using a nonlinear regression of the S-Nf data. S is the applied stress (MPa) and Nf is the number of cycles to failure. The lower asymptote of the S-Nf curve A represents the fatigue limit of the cement. The upper asymp-tote of the S- Nf curve B is the ultimate compressive strength of the cement at 1 cycle. C is the number of cycles at the inflection point of the curve and D is correlated to the slope at the inflection point. The fatigue life of a bone cement is defined as the number of cycles the material can withstand before failure, and the fatigue strength is the stress at which failure happens after a specified number of cycles Nf. The fatigue limit is the stress below which the material is consid-ered not to fail even for an infinite number of load cycles. The up and down method with a small number of samples (also re-ferred to as the Staircase method in some studies 112–114) was em-ployed in this thesis. It was originally developed for research fields dealing with sensitivity experiments such as testing of explosives, the acute toxicity of drugs, or fatigue testing of metals 115,116. This proce-dure, when applied to the field of fatigue testing, requires selecting a series of consecutive stress levels to apply to the cements (S2, S1, S0, S-1, S-2 and so on). If the specimen fails at the stress level S0, the next specimen is tested at S-1, otherwise if it survives, it is tested at S1. In this way, the applied load is adjusted up or down, depending on the outcome of the previous test. Thus, this method of experimentation helps concentrate experimental effort to the most relevant stress lev-els. This results in a large reduction in the number of specimens that need to be tested, 117,118. The starting stress level for fatigue testing (Paper III and IV) was chosen at approximately two-thirds of the qua-si-static compression strength, and steps of 2.5 MPa were used to identify the relevant stresses 66,105.

40

3.2.3 Further characterization An ideal bone cement for vertebral augmentation should be injectable (i.e. have adequate viscosity) and easy to handle. It should have high radiopacity to facilitate the tracing of its placement in the vertebral body, an adequate setting time (approx. 15 min) and a low curing temperature. Therefore, in addition to mechanical testing, some of these parameters were assessed and various analytical techniques were employed to characterize the new formulation of bone cement.

Handling properties The handling properties (Paper I, II and V), in particular the setting time (tsetting), the doughing time (tdoughing) and maximum polymeriza-tion temperature (Tmax) of the cements, were determined according to ISO 583364. Setting time is defined as the time taken to reach a tem-perature midway between ambient and maximum polymerization temperature. Doughing time is defined as the time at which a freshly dough do not stick to the surface of powder-free latex glove.

Injectability Evaluation of the injectability of the cement is required for clinical use as the cement will be injected percutaneously through cannulas into the vertebral body during a limited period of time. It gives an estimation of how long and how well a cement can flow and be manipulated. For this reason, injectability tests were performed in Papers II and V, as descibed in previous studies 119, and the force needed to extrude the cement was recorded as a function of time.

Differential scanning calorimetry (DSC) The glass transition temperature (Tg) in Paper I and II was determined via differential scanning calorimetry (DSC) and corresponds to the

41

temperature at which the polymeric cement has a transition from a hard and rigid glass material to a rubber-like and soft material. A change in the Tg of the material is related to other physical properties (stiffness, thermal expansion coefficient, etc).

Monomer release analysis The amount of monomer released during setting of the cement was measured since the unreacted MMA monomer is known to be cytotox-ic in vitro 120–122. Measurements were performed in Paper IV and V as recommended by ASTM F451 123, for monomer analysis of cured bone cement. Monomer analysis was performed by headspace gas chromatography-mass spectrometry (HS-GC/MS).

Porosity measurement Sample porosity was assessed radiographically using micro-computed tomography (microCT) at a source voltage of 100 kV and a current of 100 µA using a Cu-Al filter and quantified using the microCT soft-ware CTAn. In addition to microCT, the amount of pores on the sur-face and at the cross-section of the samples was examined via scanning electron microscopy (SEM) and CTAn software was used to quantify the amount of pores as a percentage of voids. Porosity and cross-sections of the samples were studied in Paper I, II, and III.

Pullout testing PMMA bone cement can be used to augment pedicle screws to avoid construct failure by enhancing the screws’ stability in poor quality bone 124,125. To determine the strength of the augmented construct, pullout tests were carried out (Paper II) based on ASTM F543-07 126. Cylindrical blocks of bone cement (20 x 20 mm) were prepared and half threaded cancellous screws (Titanium orthopaedic screws, HB 6.5 mm, Jiangsu IDEAL Medical Science & Technology, China) were

42

inserted in the center. The force required to remove the screw from the polymeric specimens was recorded as a function of the displacement. Similarly, pullout tests were performed in polyurethane Sawbone® cubes (blocks of 30 x 30 x 40 mm, 0.24 g.cm-3 purchased from Saw-bone® Europe AB) and the holding power of augmented screws was evaluated.

Autoclaving process Surgical interventions imply the use of sterile materials to avoid risks of infections. As mentioned previously, PMMA bone cement was provided in a ready-to-use sterilized package, while the additive lino-leic acid (LA) was supplied in a non-sterilized vial. LA was therefore autoclaved at 121°C for 20 min and the effect of sterilization was ana-lysed using proton nuclear magnetic resonance spectrometry (1H-NMR) and supercritical fluid chromatography-tandem mass spectrom-etry (SFC-MS/MS). Quantification of sterilized and non-sterilized LA was carried out by evaluation of particular peaks in 1H-NMR and SFC-MS/MS. This data can be found in Paper II.

3.2.4 Biocompatibility: in vivo evaluation The biological performance of LA-modified bone cement was as-sessed in vivo in Paper V, using an ovine model. This study was per-formed based on ISO 10993-6 standard 127. Regular and LA-modified cements were implanted in defects (cylindrical hole of 4 mm diameter and 9-12 mm length) created in the femur and humerus of sheep after complete anesthesia of the animal. Animals were observed daily until termination at 4 and 12 weeks after implantation. Bone healing was analyzed by histomorphometric analysis, which is a technique commonly used to obtain quantitative information on bone microarchitecture, remodelling and metabolism. After sample preparation and staining, the bone slides were visualized in a

43

microscope equipped with a color image analyzing system to perform a semi-automatic analysis. A standardized region of interest (ROI) ring was defined as shown in Figure 7. The proportions of the implant (cement), lacunae, fibrous tissue and bone area densities within the ROI were evaluated. Tissue inflammation was analyzed using histopathologic analysis. Bone-healing parameters were evaluated and factors such as bone remodelling or concentration of osteoblastic cells were graded from 0 to 4 (0: not detected, 1: slight evidence, 2: moderate evidence, 3: marked evidence, 4: severe evidence). A similar system of grading was employed to quantify the cellular inflammatory response (0: not detected, 1: rare, 1-5/phf, 2: 6-10/phf, 3: heavy infiltrate, 4: packed; phf = per high powered (400x) field) and factors such as concentration of osteoclasts or bone necrosis were assessed. Evaluation of the cement radiopacity and analysis of blood samples was also conducted.

Figure 7: Illustration of the region of interest (ROI) defined for histomorphometric analysis

Implant

ROI (1 mm thickness)

Bone tissue

ROI-Bone tissue interface

Initial margin of the defect of 4mm Ø

44

4. Results

4.1 Handling properties The handling properties of the bone cements were evaluated in terms of setting time (tsetting), doughing time (tdoughing), and peak temperature (Tmax) in Papers I, II and V. These data are presented in Table 7. The injectability curves were also included in this section and are shown in Figure 8. In paper I, the vertebroplastic cement Osteopal® V was modified with fatty acids (FA) and triglyceride oils (TO). As seen in Table 1, the modified cements exhibited significantly lower tsetting and Tmax. Unlike Osteopal® V cement, V-SteadyTM and F20® cements displayed (slight-ly but significantly) longer tsetting and tdoughing when linoleic acid (LA) was added to the formulation (Papers II and V). However, the Tmax of both cements decreased in a similar manner to Osteopal® V cement. Depending on the brand, the reduction of Tmax could be as high as 60%. The injectability of modified and unmodified V-SteadyTM and F20® cements was evaluated in Papers II and V. This test was performed in air, at room temperature (21± 0.5 °C). According to the manufactur-ers, the “application phase” of these cements begins at 5 min from the start of mixing at this temperature; therefore the injectability test was started 5 min after start of mixing. The cements were extruded at slow (1.5 mm.min-1) and moderate (5 mm.min-1) rates, as an optimal filling of a vertebra, besides other factors, may depend on the injection pres-sure and rate used during surgery128–131. The test was stopped when a force of 150 N was reached, since this limit was reported to be ap-proaching the physical limits that a clinician could apply during verte-broplasty131,78.

45

The use of LA (sterilized and non-sterilized) may have an effect on the injectability of regular cements. Depending on the brand, the mod-ified cements could be injected for similar (F20®-LA cement) or long-er times (V-SteadyTM-LA cement) than the control cement, as seen in Figure 1. The modified cements could be extruded at low (1.5 mm.min-1) and high speed (5 mm.min-1) until reaching the limit of 150 N, and was still manually injectable after the end of the test. The in-jection speed had no significant effect on the injectability (in terms of % injected material) of these cements and the LA-modified cements could be injected over time without any observed phase separation. The unmodified cement could usually be injected at both high and low speeds. In particular, F20® cement was successfully injected at both speeds when the test was started at more than 5 min from the start of mixing. Conversely, V-SteadyTM cement was usually fully extruded at high speed only and its injectability (in terms of % injected material) gradually decreased when the injection speed of injection was re-duced, and the cement hardened. All control cements could also be injected without any phase separation observed.

46

Table 7: Effect of additives on the handling properties of acrylic bone cements. OP = Osteopal® V cement; VS = V-SteadyTM cement; F20 = F20® cement; LA= linoleic acid; RA = ricinoleic acid; CO = castor oil. The ‘ * ’ indicates that the LA had been sterilized; ‘†’ indicates wt% of the total cement; ‘††’ indicates vol% of the monomer phase. The Tg was determined at 24h for all cements, those measured at 2 weeks are indicated in parenthesis.

Cement Amount of additives tsetting [min] tdoughing [min] Tmax [°C] Tg [°C]

OP - 21.1 ± 1.1 - 63.5 ± 2.6 98.4 OPL-A 1† 19.7 ± 0.9 - 35.3 ± 0.4 86.4

OP-RA 12† 13.7 ± 0.7 - 35.0 ± 0.8 57.7

OP-CO 12.6† 15.6 ± 0.1 - 34.2 ± 2.7 74.0 VS - 22.1 ± 1.0 9.1 ± 0.4 66.8 ± 2.7 102.8 (114.0)

VS-LA 12 †† 24.7 ± 1.0 12.3 ± 0.3 28.2 ± 0.4 74.7 (87.8)

VS-LA* 12 †† 20.8 ± 1.0 9.9 ± 0.1 31.1 ± 1.1 78.0 (92.8)

F20 - 54.3 ± 2.0 - 38.7 ± 1.0 - F20-LA 0.5† 59.2 ± 2.0 - 27.1 ± 1.0 -

47

Figure 8: Injectability curves of regular V-SteadyTM cement, V-SteadyTM-LA cement (Non-Sterilized = VS-LA (NS); Sterilized = VS-LA (S)), F20® cement and F20®-LA cement when injected at A) 5 mm.min-1 and B) 1.5 mm.min-1. n=3 for each group. One representa-tive specimen per group and injection speed is shown in the figure.

0

25

50

75

100

125

150

0 5 10 15 20 25 30 35 40 45

Forc

e (N

)

Time (min)

VS VS-LA (NS) VS-LA (S) F20 F20-LA

1.5 mm.min-1

!"

0

25

50

75

100

125

150

0 2 4 6 8 10 12 14 16 18 20

Forc

e (N

)

Time (min)

VS VS-LA (NS) VS-LA (S) F20 F20-LA

5 mm.min-1

!"

48

4.2 DSC analysis The glass transition temperature (Tg) of OP and VS cements are pre-sented in Table 7. The modified cements exhibited significantly lower Tg both at 24h and 2 weeks. The Tg determined at 2 weeks was signif-icantly higher than that measured at 24h for both modified and un-modified cements.

4.3 Quasi-static mechanical properties The quasi-static mechanical properties of the cements are presented in Figures 9, 10, 11, 12 and 13. In Paper I, different formulations of acrylic cements containing FA and TO were prepared. The mechanical properties of Osteopal® V cement was affected to a similar extent by the addition of additives, with a significant decrease in strength and stiffness of up to 75-85%, as seen in Figure 9. Figure 10 shows the stiffness and strength over time of V-SteadyTM and F20® cement, compared to a commercially available low-modulus acrylic bone cement, Resilience®. Although there are slight variations (not statistically significant), the unmodified cements VS and F20 exhibited similar mechanical properties over time, comparable to oth-er commercially available standard acrylic cements. The stiffness and strength of the modified cements were reduced in average by almost 71% and 65%, respectively, after 24h. However, the mechanical prop-erties of both cements increased progressively over time, whereas, as also stated by the manufacturer, the stiffness and strength of Resili-ence® decreased successively over the first 2 weeks. The compressive mechanical properties of V-SteadyTM and V-Steady-TM-LA cements were also evaluated in PBS at 37°C after setting for 2 weeks under the same conditions. This test revealed that, compared to the modulus in air, the modulus of unmodified cement significantly increased when tested in PBS (from 2075 ± 114 MPa to 2675 ± 102

49

MPa). However, the stiffness of the modified cement remained similar when tested in air and in PBS. Conversely, the ultimate compressive strength (UCS) of both cements significantly decreased with a loss of strength of 12 MPa (from 96 ± 5 MPa to 84 ± 4 MPa) and 10 MPa (31 ± 1 MPa to 21 ± 1 MPa) for VS and VS-LA respectively, as seen in Figure 11. In Paper II, the flexural properties and pullout force of V-SteadyTM cement modified with LA were determined. There as well, the modi-fied cement displayed significantly lower mechanical properties (Fig-ures 12 and 13). In particular, when submitted to bending tests, V-SteadyTM cement exhibited high stiffness (bending modulus, Eb, rang-ing between 2900 and 3200 MPa) and could be loaded until fracture both after 24h and 6 weeks of setting. Unlike the unmodified cement, V-Steady-LATM cement showed high flexibility and bent without frac-turing until week 6 (Figure 12). The use of LA allowed a decrease of its Eb by almost half (Eb = 1565.7 MPa). A similar trend was seen in the pullout force of screws from modified cements tested in Sawbone®. During this test, the use of LA led to a reduction of the holding strength compared to the regular cement of up to 72%.

50

Figure 9: Compressive elastic modulus (E modulus) and ultimate compressive strength (UCS) of Osteopal® V cement without any addi-tive and prepared with 1 wt% linoleic acid (LA; n=6), with 13.8 wt% ricinoleic acid (RA; n=6) and with 12.8 wt% castor oil (CO; n=8).

0

10

20

30

40

50

60

70

80

90

0

200

400

600

800

1000

1200

1400

1600

1800

OP OP-LA OP-RA OP-CO

UC

S (MPa)

E m

odul

us (M

Pa)

E modulus UCS

Modulus of vertebral trabecular bone

Strength of vertebral trabecular bone

51

Figure 10: Mechanical properties (E modulus and ultimate compres-sive strength) of regular and low-modulus bone cements over time in 37C PBS. VS-LA contained 12vol% of linoleic acid, the ‘ * ’ indicates that the LA had been sterilized (VS-LA*). F20-LA was prepared with 0.5wt% linoleic acid. For all groups, n=6.

0

500

1000

1500

2000

2500

VS VS-LA VS-LA* F20 F20-LA Resilience

E m

odul

us (M

Pa)

24h 2 weeks 4 weeks Modulus of vertebral trabecular bone

0

20

40

60

80

100

120

VS VS-LA VS-LA* F20 F20-LA Resilience

UC

S (M

Pa)

24h 2 weeks 4 weeks Strength of vertebral trabecular bone

52

Figure 11: Compressive elastic modulus (E modulus) and ultimate compressive strength (UCS) of regular cement (V-SteadyTM, VS) and low-modulus cement containing 12vol% of linoleic acid (V-SteadyTM-LA, VS-LA) after storage in PBS for 2 weeks, tested in compression either in air at room temperature or in PBS (37°C). For all groups, n=6.

0

20

40

60

80

100

120

VS VS-LA

UC

S (M

Pa)

2 weeks, Air 2 weeks, PBS 0

500

1000

1500

2000

2500

3000

VS VS-LA

E m

odul

us (M

Pa)

2 weeks, Air 2 weeks, PBS

53

Figure 12: Flexural properties of V-SteadyTM cement (VS) and V-SteadyTM containing 12vol% of linoleic acid (VS-LA) at 24h and 6 weeks. The modi-fied cement did not break break at the 24h. For all groups, n=6.

0 10 20 30 40 50 60 70

VS VS-LA

Ben

ding

stre

ngth

(MPa

)

24h 6 weeks

0 500

1000 1500 2000 2500 3000 3500 4000

VS VS-LA

Ben

ding

mod

ulus

(MPa

)

24h 6 weeks

*

54

Figure 13: Pullout force of V-SteadyTM cement (VS) and V-SteadyTM-LA cement (VSLA) augmented screws from Cement only (pilot hole of 3.8 mm; n=5) and from Sawbone® (pilot hole of 3.2 mm; n=8). The pull-out force from Sawbone® only is 36.3 ± 8.4 N 132.

4.4 Compressive fatigue properties The fatigue properties in compression of regular and low-modulus bone cements were investigated in Papers III and IV. Figure 14 shows the S-Nf curves and fit to the Olgive equation. The material constants determined from the Olgive equation are summarised in Table 8. In Paper III, the fatigue analysis was conducted in air at room temper-ature. Using the up and down method, the fatigue limit of F20® ce-ment was estimated at between 45 and 47.5 MPa since 4 specimens survived compressive loads at 45 MPa until runout, and 1 specimen survived compressive loads at 47.5 MPa. Based on a similar reason-ing, the fatigue limit of Resilience® was estimated to be approx. 30 MPa (3 specimens survived loads at 30 MPa). The fatigue limits of these cements were also determined from the S-Nf curves obtained from the Olgive equation fits (Figure 14-A) and

0

1000

2000

3000

4000

5000

6000

7000

8000

VS VS-LA

Pullo

ut F

orce

(N)

Cement only

Sawbone + 1 ml cement

55

were estimated at 47 MPa for F20® and 31 MPa for Resilience® (Pa-rameter A in Table 8). In Paper IV, the fatigue tests were performed in a biobath containing PBS solution at 37°C. The fatigue limit of the V-SteadyTM cement was estimated to be between 42.5 and 45 MPa since 5 V-SteadyTM speci-mens reached the runout of 2 million cycles at these 2 compressive stress levels with the up and down method. Similarly, the fatigue limit of the modified cement was estimated at between 2.5 and 5 MPa. Ad-ditional tests performed midway at 3.75 MPa confirmed this hypothe-sis, and further narrowed the range to between 3.75 and 5 MPa, as all 3 specimens survived at this stress level. The Olgive fit gave a fatigue limit of 4.7 MPa for the modified cement (Figure 14-B; Parameter A in Table 8). However, the Olgive equation did not fit the data of the unmodified cement V-SteadyTM, and no fa-tigue limit could be estimated with the equation for this cement. Similarly to the quasi-static mechanical properties, LA affected the fatigue performance of the V-SteadyTM cement and considerably re-duced the fatigue strength of the standard acrylic cement by approx. 90%. Table 8: Estimated Olgive parameters for regular (F20®) and low-modulus cements (Resilience, V-SteadyTM-LA). The 95% confidence limits are indicated in parenthesis. A = lower asymptote (fatigue lim-it); B = upper asymptote; C = number of cycles at the inflection point of the curve; D = slope at the inflection point.

A B C D F20® 47.0 75.0 2.2 7.7

(45.4; 48.6) (70.5; 79.4) (2.0; 2.3) (4.5; 10.8) Resilience® 30.9 51.6 2.5 3.8

(28.9; 32.8) (48.4; 54.7) (2.2; 2.9) (1.6; 6.0) VS-LA 4.7 22.8 2.7 16.5

(3.8; 5.7) (20.3; 25.4) (2.6; 2.8) (5.2; 27.7)

56

Figure 14: S-Nf curves of regular (F20®, V-SteadyTM / VS) and low-modulus cement (Resilience®, V-SteadyTM-LA / VS-LA). The dashed curves are the 95% confidence limits. Fatigue testing was performed A) in air and B) in PBS at 37°C.

4.5 Additional material characterization

4.5.1 1H-NMR and SFC-MS/MS analysis The effect of autoclaving on the LA compound was investigated to determine whether heat treatment affected its composition. 1H-NMR analysis revealed that there were no differences in composition for the LA compound before and after autoclaving, whereas the SFC-MS/MS analysis showed that approx. 18% of the LA had degraded after auto-claving. For both techniques, no additional peaks were detected.

Experimental data – F20 Experimental data – Resilience Olgive model – F20 Olgive model – Resilience

Stre

ss (M

Pa)

90 80 70 60 50 40 30 20 10 0

Experimental data – VS Experimental data – VS-LA Olgive model – VS-LA

Stre

ss (M

Pa)

90 80 70 60 50 40 30 20 10 0

0 1 2 3 4 5 6 7 0 1 2 3 4 5 6 7 Log(Nf) Log(Nf)

57

4.5.2 Monomer release In Paper IV, the amount of unreacted monomer released by the low-modulus commercially available bone cement Resilience® was com-pared with the release from the unmodified V-SteadyTM and the exper-imental cement V-SteadyTM-LA. As seen in Figure 7, the regular ce-ment V-SteadyTM released relatively low amounts of unreacted mon-omer compared to the low-modulus cements V-SteadyTM-LA and Re-silience®. The addition of LA into the V-SteadyTM formulation resulted in an increase of unreacted monomer. The V-SteadyTM-LA cement liberated 5-fold more unreacted MMA than the control during the first hour and continued to release more monomer over time. However, this cement liberated less or comparable amounts of mono-mer to the commercially available Resilience® cement over time as shown in Figure 15. In particular, it liberated 2-fold less unreacted monomer during the first hour of polymerization. The cements Osteopal® V and F20®, which were evaluated in Papers I and V respectively, released much more unreacted monomer than the previous brands of acrylic cements (14 to 16-fold more compared to V-SteadyTM cement at 24h). Similarly, the low-modulus formulations of both cements released significantly higher amounts of monomer than the regular cements Osteopal® V and F20®. In summary, the addition of linoleic acid resulted in an increase of unreacted monomer release for all cements used.

58

Figure 15: Concentration of unreacted MMA monomer from cured control (F20®, Osteopal® V, V-SteadyTM) and low-modulus bone ce-ments (Resilience®, F20®-LA, Osteopal® V-LA, V-SteadyTM-LA); n=12 (per group and time points) for all groups of cement except for OP, where n = 6 (per group and time points).

4.5.3 Porosity The influence of porosity on the dynamic properties of acrylic bone cements was investigated in Paper III. In this work, the porosity of F20® and Resilience® cements was first assessed using microCT measurements (Table 9 and Figure 16). These results were then com-pared to data obtained by CTAn software analysis of SEM images taken of the specimens’ surface as well as bulk cross-sections (Table 9). Table 9 shows that the porosity of F20® cement was relatively low (between 1.4 – 7%) and remained constant over time, even though the microCT and the SEM images gave different percentages of porosity.

0

400

800

1200

1600

2000

2400

2800

0 1 2 3 4 5 6 7 8

MM

A (m

g/L)

Time (days)

Resilience VS VS-LA F20 F20-LA OP OP-LA

59

According to both analytical techniques, Resilience® was more porous than F20®. As stated by the manufacturer, the porosity of Resilience®

cements increases over time due to poly(amino acid) (PAA) particles that are dissolving. This was however not detected by the microCT analysis since the low radiopacity of the PAA particles in comparison with the polymer matrix containing high amounts of radiopacifier made the PAA particles invisible to this technique. Hence the varia-tion in porosity could only be demonstrated by using the SEM images of the specimens’ cross-sections (Table 9 and Figure 17). There was almost no porosity at the cross section of this cement (1.4%) com-pared to its surface (9.2%) after 24h setting in PBS (Table 9). Howev-er, the porosity of the whole cement increased over time. Analysis of sample surfaces of cements modified with LA by SEM attested that the addition of LA had no effect on the microstructure of the cement. Regular and low-modulus cements exhibited similar mi-crostructures, with a polymeric matrix containing pores and radiopaci-fying agents throughout the surface. Table 9: Porosity of F20® and Resilience® cements as estimated from SEM images and microCT. n = 3 for all groups except for * n=2 and ** n =1.

Porosity (%)

SEM (2D) MicroCT (3D)

Time point Location Mean (±SD) Mean (±SD)

F20® 24h Surface 1.4 (± 1.3) 7.0 (± 2.9)

4weeks Surface 1.6 (± 1.8) 6.8 (± 3.0)

Resilience® 24h Cross-section 1.4 (± 0.4)

24h Surface 9.2 (± 2.3) 21.5 (± 0.5)

4 weeks Cross-section 4.3 (±4.0) 22.8 (± 1.0)

4 weeks Surface 16.4 (±0.0)

60

Figure 16: MicroCT images of F20® and Resilience® cements after different times of storage in PBS at 37C.

Figure 17: SEM micrographs showing the surfaces of A) F20® ce-ment, B) VS cement, C) Resilience® cement, D) F20®-LA cement and E) V-SteadyTM-LA cement after setting 24h in PBS (37°C) at magnifi-

300 μm

A

(d) (f)

300 μm

300 μm

B

D E

300 μm

F

300 μm300 μm

300 μm

C

300 μm500 μm

500 μm

F20®

4 weeks 24h

Resilience®

2 μm

61

cation x100. Pictures F) shows the cross-section of Resilience® ce-ment. The inserts in image C) and F) show the surface and cross-section (respectively) of Resilience® after 4 weeks setting in PBS at magnification x100. The white dots are the radiopacifier and the black cavities are the porosity resulting from mixing, polymerization or dis-solution of PAA in the case of Resilience® cement.

4.6 In vivo evaluation In Paper V, F20® cement and its low-modulus formulation containing 0.5 wt% LA were implanted into the femur and humerus of sheep to this end. Blood analysis revealed that the implantation of F20®-LA cements in the animal did not disrupt the normal content of fatty FA of the blood. The FA components in sheep implanted with the modified cement remained similar over time (despite slight variations due to fasting) and comparable to those of animals implanted with regular cements. The radiopacity of both cements (unmodified and modified) was established. As seen in Figure 18, both cements displayed an intense radiopacity at the implanted sites, indicating that the addition of LA into the formulation of regular cement had no effect on the radiopacity. Histopathologic and histomorphometric analysis attested that the use of LA did not compromise the biocompatibility of regular cement. In particular, there was no evidence of cytotoxic effects, and both regular and F20®-LA cements exhibited similar inflammatory responses over time. A thin fibrous membrane formed around both cements and was visible 4 weeks after im-plantation (Figures 19 and 20). The thickness of the fibrous layer decreased by approx. 65-70% at week 12, as shown in Figure 20. Both materials showed an active bone healing process, with the presence of osteoblasts at the bone-cement interface. The inflammatory response was progressively attenuated and new bone was formed over time while bone remodelling oc-curred, as indicated by the slight increase of bone marrow in Figure 20. The in vivo evaluation of LA-modified cement demonstrated that the implanta-tion of this cement results in a harmless tissue reaction in sheep, comparable to that observed with regular PMMA bone cement.

62

Figure 18: X-ray images of F20®-LA cement implanted in A) right humerus and B) left femur (sheep 1); and F20® cement implanted in C) right humerus and D) right femur (sheep 3) at termination 4 weeks.

Figure 19: Histological cross-sections of F20® and F20®-LA speci-mens implanted in the femur and humerus of sheep. The photomicro-graphs show A) F20-LA cement and B) F20® cement at 4 weeks and C) F20®-LA cement and D) F20® cement at 12 weeks. The sections were stained with modified Paragon. Bone tissue BT (pink), radiopac-

100 μm

F T

B T

C R

C

R

B T

B M

C

100 μm 100 μm C R

F T

B T

R

C

B T

B M D

100 μm

A B C D

20 mm

63

ifier R (black dots), Cement C (white), fibrous tissue FT (navy blue), Bone marrow BM.

Figure 20: Proportions of fibrous tissue and bone area determined with histomorphometric analysis.

0

10

20

30

40

50

60

F20-LA F20 F20-LA F20

Perc

enta

ge (%

)

Fibrous tissue area Bone marrow area

64

5. Discussion

The aim of this thesis was to develop and evaluate a PMMA-based bone cement with bone-like mechanical properties. This work started with the evaluation of different brands of cements containing varying amounts of additives, and the evaluation of some of their functional properties. This was reported in Paper I, where the addition of fatty acids (FA) and triglyceride oils (TO) positively affected the handling and quasi-static mechanical properties of regular PMMA cements. FAs, and in particular linoleic acid (LA), appeared to have a greater effect on the final mechanical properties of the cement compared to TOs, even when low amounts FAs were used. This might be related to the difference in level of unsaturation of FAs and TOs, commonly measured in terms of iodine value (mass of iodine in grams that is consumed by 100 g of a substance). Indeed, LA has a higher iodine value (181.0 mg) than RA (85.1 mg), which probably makes it more reactive within the bone cement. In addition, the influence of the io-dine value might be enhanced by the increased availability of the end hydrogen atom for abstraction. Therefore, the preparation of LA-modified cements became the main focus of this thesis, supported by promising in vitro results 86,133. Cement properties were altered because the additives, and in particular LA, were likely to give rise to a combination of effects. Different mechanisms have been suggested and described in the literature. For instance, Guo et al134 suggested that LA chemically modifies PMMA during polymerization. Indeed, methacrylation of LA is likely to occur by grafting of the free methyl methacrylate (MMA) radicals onto the LA chains via abstraction of the alpha hydrogen. This results in the formation of stable radical sites on the fatty acid, and therefore causes a reduction of the polymerization rate as the polymeric chains shorten and the molecular weight of the matrix decreases134. Consequently, a

65

pronounced drop in E modulus is observed, accompanied by a reduc-tion of the strength and alterations of the handling properties. LA can also act as a natural based plasticizer135,136 and improve the flexibility and processability of polymers by lowering their glass tran-sition temperature (Tg) and E modulus79,135,136. The plasticization pro-cess, occurring when the molecules of LA are introduced into the pol-ymer matrix, might affect the material through different mechanisms. First, the LA molecules could create additional free volume to the polymer chain segments and allow internal chain rotation and mobili-ty135,137. Besides this, they could also act as a lubricant and reduce the internal friction between the polymer chains137. Moreover, Vallo et al.138 reported that the residual monomer MMA might act as a plasti-cizer, and that an increased amount of MMA due to the presence of additives might contribute to a change in properties. Interestingly, the evaluation of the mechanical properties over time (in compression and in bending) revealed that, contrary to regular cement, the stiffness and strength of LA-modified cements (F20®-LA and V-SteadyTM-LA) increased over time (Figure 10). This could be associ-ated to washing out of the LA and MMA monomer 138 and slow mo-lecular motions within the material progressively reducing the free volume, densifying the material and increasing the mechanical proper-ties. While the cements continue to harden over time, the increase in mechanical properties is accompanied with an increase of the Tg, as confirmed by the higher value of Tg after 2 weeks (Table 7). Conversely, the stiffness and strength of Resilience® cement succes-sively decreased over time and this was attributed to the presence of the poly(amino acid) (PAA) component in the cement formulation. PAA is a biodegradable polymer that dissolves in vivo or in an aque-ous solution and causes porosity formation in the PMMA bone cement matrix. A similar outcome was demonstrated by Puska et al.77, who evaluated the mechanical properties of Palacos® R cement containing 20 wt% of an oligomer filler based on an amino acid of trans-4-hydroxy-L-proline. In Paper I, Osteopal® V modified with FA and TO, took less time to set compared to the regular cement while F20® and V-SteadyTM modi-fied cements required longer time (compared to the regular cement).

66

The divergence in the outcome was related to the different methods used to prepare the cement. In the first study, the modified cements were prepared by removing 54% of the additive volume from the monomer phase. Therefore, the initiator was present in excess and a smaller amount of monomer was available to react with the initiator, preventing the formation of long polymeric chains and resulting in shorter setting time. In the other studies, LA was directly introduced in the liquid phase. The number of molecules of initiator (in excess in this formulation) and activator being unchanged, the presence of LA did not compromise to a large extent the formation of the radicals ini-tiating the polymerization reaction, which resulted in relatively longer setting time 139. The use of LA caused a decrease in the maximum polymerization temperature of regular cement, likely due to the reduction of the polymerization rate. From a clinical point of view, a reduction of Tmax is positive, as it would benefit the surrounding bone tissue directly exposed to the cement 140–142. Indeed, it has been reported that temper-atures as high as approx. 120°C could be generated by large amounts (>25g) of cement tested in vitro143,144 as described by the ASTM standard123. However, it is unlikely that such temperatures occur in vivo. The amount of cement generally injected during vertebroplasty (<10ml) is lower than the one required by the standard for measuring the setting time and temperature46, and recent studies have shown that the use of lower amounts of cement (≈ 2ml) is enough to stiffen and stabilize the vertebrae, as well as reduce back pain 46. This would re-duce the risk of thermal necrosis of tissues next to the vertebra, as a lower amount of cement would generate a lower temperature 145. Despite the positive effects on the mechanical and handling properties of regular cements, the addition of LA caused a negative change in monomer release. LA-modified cement was shown to liberate larger amounts of unreacted monomer. This outcome appears to be highly dependent on the cement formulation, since F20® and Osteopal® V cement, which contained 0.5 wt% (≈ 2vol%) and 1 wt% LA respec-tively, released higher concentrations of MMA than VS cement with 12 vol% LA. MMA cytotoxicity is a general concern and has been associated with bone necrosis of the vertebral bone tissue146.

67

In a study by Lopez et al.86, Osteopal® V with 1.5 wt% LA also re-leased a high amount of MMA during the first 6 hours of incubation. Furthermore, in the in vitro cytotocixity assay, the same cement showed a lower number of viable cells under undiluted conditions. However, at a 4-fold dilution, this effect was no longer observed86,133. Therefore, it was assumed that in the in vivo situation, the small vol-ume of cement used to fill the defect would release very small quanti-ties of unreacted monomer, which would be progressively diluted in the body fluid. This hypothesis was corroborated in the in vivo study (Paper V) included in this thesis. Indeed, no adverse effects related to either thermal or chemical necrosis were noticed after evaluation of the region of interest of the bone defect. In addition, this result was obtained despite choosing a worst-case scenario, with the slow-setting base cement F20®, which released the highest amounts of MMA in vitro (up to 2419 µg/mL at 24h, Figure 15) of all the cements in this thesis. Currently, the limiting amount of free monomer released from acrylic bone cement has not been fixed to a specific concentration. An in vitro study showed that an MMA blood concentration of up to 5-10 µg/mL was cytotoxic to blood cells and endothelial cells, even during a short exposure (1 min)120. These doses were however lower than some val-ues reported during routine total hip replacement, where the MMA blood concentrations varied between 0.02-59 µg/mL shortly after ap-plication of the bone cement (1-2 min)147. Despite these high levels of MMA, no pharmacological effect could be demonstrated147. Cement injectability was not negatively affected by introducing LA into regular cement. On the contrary, the modified cements demon-strated longer injection times (or application phases), which is gener-ally more convenient for the surgeon manipulating the implant. This was clear in the in vivo study, Paper V, during implantation of the ce-ment into bone defects of sheep. During in vitro analysis, the cements were injected at moderate and low injection rates to imitate cement injection during surgery. It has been reported that injecting cement during vertebroplasty may be difficult due to the high pressure re-quired (up to 1500 kPa), which can result in insufficient or poor filling of the vertebra129,130. Factors such as cement viscosity, injection rate

68

and pressure, needle size and geometry148, as well as the time elapsed between the cement preparation and injection, are all important for ensuring an optimal bone filling. Of these, cement viscosity is of ut-most importance, as cements with lower viscosities may require lower injection pressures, but entail a higher risk of leakage. In addition, the permeability of the vertebral trabecular bone structure varies from one patient to another. Clinicians have attempted to counteract problems related to cement delivery by increasing the recommended monomer-to-powder ratio, which decreases the viscosity of the material and pro-longs the injection time149,150. However, as mentioned above, besides changing the physical properties of the cement, this may lead to a higher risk of cement extravasation. In the present work, cement viscosity was not evaluated but it was established from the injectability experiments that cements containing LA resulted in an increased working time. Bohner et al.131 recom-mended a reduction in the injection rate and favoured the use of nar-row syringes for the injection of low viscosity cement. This would give better control over the injection flow and reduce the risk of ce-ment leakage. Otherwise, the use of high viscosity cement could be an option. However, Breusch et al.151 reported that such a cement may increase the risk of fat embolism by pushing away the bone marrow. The same authors observed that high viscosity cement penetrated bet-ter into the cancellous bone than low viscosity cement151, contrary to expectations. Determining an adequate viscosity to avoid the problems discussed above has proved challenging. Commercially available vertebroplastic bone cement generally con-tains high amounts of radiopacifier in the powder (between 30-45wt%), to facilitate cement visualization during injection under fluoroscopy guidance152. Zirconium dioxide (ZrO2), and barium sul-phate (BaSO4) are the most common contrast agents, even though tan-talum and tungsten powder have also been used42,44,146. F20® cement, which contains 45 wt% ZrO2, is the only bone cement whose radio-pacity has been evaluated in this thesis, in the in vivo study (Paper V). The modified cement, F20®-LA exhibited intensive radiopacity, com-parable to that observed with the regular cement. This indicated that

69

the use of small quantities of LA did not affect the radiological prop-erties of the modified cements. Similarly to the unmodified cement, a thin fibrous connective membrane was formed around F20®-LA cement implanted in vivo (Paper V). This is a typical response to PMMA cements after implantation in bone tissues153–157. There was no evidence of local cytotoxicity or bone necrosis and the modified cement showed minimal adverse reactions at the bone/implant interface, as demonstrated by the histological and histomorphometry analysis. This study appeared to confirm that the addition of small amounts of LA to PMMA bone cement is safe, since it does not cause additional cytotoxic effects to the tissue surrounding the implantation sites. This was in agreement with the findings of previous in vitro and in vivo studies using PMMA bone cement modified with LA133. As seen in Papers I and II, standard cement and LA-modified cement displayed different quasi-static mechanical properties. This difference is likely to influence their properties under repeated loading, in partic-ular under in vivo conditions. In this thesis, two studies were dedicated to the long-term mechanical performance of standard and low-modulus cement, in air at room temperature (Paper III) and in PBS solution at 37°C (Paper IV). To the best of the author’s knowledge, no other study has reported on the mechanical properties of low-modulus cements over time in vivo or in simulated body fluid. In Paper III, the standard cement F20® displayed higher fatigue limit (47 MPa) than the commercially available low-modulus cement Resilience® (31 MPa). This was expected as the regular cement had higher quasi-static mechanical properties. Both cements experienced thermal softening at the highest loads, despite the low frequency used (2 Hz). However, F20® cement appeared to be more affected by the thermal softening effect, since a slight decrease of the maximum applied load caused a radical increase of the specimen life (slope of the S-N curves, Figure 14). The thermal softening effect was less pronounced in Resilience® cement, as seen from the slope of its S-N curve, Figure 14. The fa-tigue life of this cement increased progressively as the applied load was reduced. This was associated with the presence and dissolution process of the PAA compound included in the formulation. Indeed,

70

the water molecules entrapped in the matrix during setting in PBS and the resulting porosity might reduce the thermal softening effect by allowing for better heat dissipation. The PAA components and the evolution of the porosity within the matrix were also observed in the SEM pictures (Figure 17). The fatigue strength of V-SteadyTM and V-SteadyTM-LA cements fol-lowed the same trend in saline solution at 37°C (Paper IV), and V-SteadyTM cement displayed a higher fatigue limit (42.5-45 MPa) than LA-modified cement (3.75-5 MPa) according to the up and down method. The Olgive equation did not fit the data of V-SteadyTM ce-ment, which appeared to survive a higher number of cycles than ex-pected at relatively high loads. Further investigations are required to describe the long-term performance of V-SteadyTM cement, possibly using a higher number of cycles to run-out. In Paper IV as well, the cements were affected by thermal softening (the generation of heat within the material) under dynamic cyclical loading under physiological conditions. The quasi-static mechanical properties of these cements when compressed in air or in physiologi-cal conditions differed to some extent (Figure 11). In particular, V-SteadyTM cement showed a significant increase in stiffness and a slight (but not statistical) decrease in strength in PBS. It was not clear why the stiffness of the cement increased in PBS, as it was rather expected to decrease158. However, the strength of both cements (V-SteadyTM and V-SteadyTM-LA) slightly decreased in PBS, in agreement with earlier reports158. The decrease in strength was attributed to the plasti-cizing effect of water uptake. The fatigue limits exhibited by the standard and Resilience® cements were considerably higher than the intervertebral stresses measured in a vertebral body (1-3 MPa)12,159,160. V-SteadyTM-LA cement was the only cement whose fatigue strength was in the same order of magni-tude as the stresses measured in vivo. In respect to the results of quasi-static mechanical properties over time, it can be assumed that the fatigue strength of LA-modified ce-ment will follow a similar trend to that of its quasi-static mechanical properties, and its fatigue strength will increase over time. A similar reasoning could imply that the fatigue strength of Resilience® cement

71

may decrease with time. In the case of osteoporotic patients with VCFs, it is debatable whether it is more convenient to use low-modulus cements exhibiting either immediate (LA-modified cement) or progressive (Resilience® cement) bone-compliant mechanical prop-erties. The majority of the adjacent fractures - which account for 36-67%50–54 of the additional fractures - occur between 1 to 4 months after treatment50,53,161–163. Therefore, the use of a cement displaying lower modulus immediately after implantation could reduce the risk of additional VCFs. Resilience® cement - which has mechanical proper-ties in agreement with the requirements of ASTM standard after 24h of setting101- exhibited mechanical properties approaching those of trabecular bone only after some weeks in PBS, and these seemed to plateau between 2-4 weeks as seen in Paper III. Besides potential use in the treatment of osteoporosis-induced vertebral compression fractures, LA-modified cement might also be appropriate for use in applications where hardware needs to be applied. Indeed, it showed pullout strengths comparable to the holding power of some CaP cements132,164. The benefits of regular PMMA bone cement to augment pedicle screws are well-known124,125,165, however the use of low-modulus cement might be more appropriate for the treatment of severely osteoporotic bone, although further in-vestigations are needed to confirm this. Sterilization through autoclaving appeared to be a simple and ade-quate technique to sterilize LA, as it did not significantly modify the composition of the compound. H-NMR analysis revealed that there were no differences between the LA compound before and after autoclaving, whereas SFC-MS/MS showed that approx. 18% of the LA degraded. This alteration however did not seem to hinder its ability to reduce the stiffness of regular ce-ment, and no significant variations of the functional properties of LA-modified cement (handling, quasi-static mechanical properties and injectability) were noticed. In summary, it was demonstrated in this thesis that LA-modified ce-ment is a promising candidate for the treatment of osteoporosis-induced vertebral compression fractures. Besides displaying bone-like mechanical properties, it demonstrated a harmless tissue response in

72

vivo and this, despite the general inconvenience related to the implantation of PMMA material in vivo.

73

6. Concluding remarks and Future perspectives

The work presented in this thesis demonstrates that LA-modified bone cements may be a beneficial alternative to the currently commercially available PMMA cement, for the treatment of osteoporosis-induced vertebral compression fractures. The addition of LA into regular PMMA cement caused an immediate and considerable reduction of the mechancial properties, in particular the stiffness. The amount of LA required to obtain mechanical properties that match those of human vertebral trabecular bone is dependent on the base cement’s formulation. The use of LA led to changes in handling properties and injectability of the paste. In particular, LA decreased the polymerization tempera-ture of the cement as well as increased its injection time. However, its use resulted in higher amounts of residual monomer released during the polymerization reaction. This drawback did not appear to signifi-cantly affect the in vivo performance of the material, which showed a biocompatibility comparable to that observed with regular PMMA bone cement. The decrease in quasi-static mechanical properties im-plied a decrease also of the fatigue limit of the modified cements, to stresses of the same order of magnitude as those measured in vivo, although still higher. Many topics remain to be investigated in the field of low-modulus cement designed for the treatment of osteoporotic patients. In particu-lar, the long-term behavior of low-modulus cement with longer stor-age times in physiological solutions needs to be evaluated, since the quasi-static mechanical properties changed over time. This would help to better correlate quasi-static and dynamic mechanical properties as well as to predict the in vivo performance of these cements over time. Further in vivo work with PMMA-based bone cements modified with LA is likely needed since the extent to which the mechanical proper-

74

ties are affected depends on the amount of LA in the formulations, which will differ depending on the base formulation. Finally, the use of low-modulus PMMA bone cement could be benefi-cial to restore the stiffness of the treated vertebra to pre-damage levels 80,82. However, the volume of cement required for sufficient rein-forcement while at the same time reducing the risk of cement extrava-sation is yet to be determined. Table 10: Summary of the main functional properties of VS cement evaluated in this thesis. The results presented here are those obtained at 24h, except when indicated by ‘*’ (6 weeks). The pullout forces are the data obtained from pullout test in augmented Sawbone®.

Regular PMMA

cement LA-modified

cement

Elastic modulus (MPa) 2140.4 494.7

Ultimate compressive strength (MPa) 100.7 28.3

Bending modulus (MPa)* 2903.7 1565.7

Bending strength (MPa)* 50.1 23.5

Pull out force (N) 2788.4 777.4

Peak Temp. (°C) 66.8 28.2

Setting time (min) 22 25

Doughing time (min) 9 12

Working time (min) 8-13 16-21

MMA release (mg/L) 98.9 870.8

Biocompatibility Yes Yes

75

Svensk Sammanfattning

Förbättrade levnadsvillkor och bättre vård har lett till en befolkning med ökad livslängd. Med den högre andelen äldre ökar även antalet personer som lider av benskörhet (osteoporos) med efterföljande kot-kompressionsfrakturer. Över hela världen finns det mer än hundra miljoner äldre som lider av benskörhet och varje år rapporteras cirka 520 000–750 000 kotfrakturer i Europa respektive USA. Kotfrakturer behandlas oftast konservativt, men för svåra frakturer med långvarig smärta kan minimalt invasiva kirurgiska behandlingar, såsom vertebroplastik (VP) och ballongkyfoplastik (BKP), användas för att stabilisera frakturen och lindra smärtan. VP och BKP består av en perkutan (genom huden) injektion av ett bencement, vanligtvis akrylplasten polymetylmetakrylat (PMMA), in i den trasiga kotan un-der röntgengenomlysning. Båda behandlingarna anses vara säkra och ger som regel omedelbar smärtlindring i 80-90% av fallen. Dessa in-grepp har dock orsakat en viss kontrovers på grund av de komplikat-ioner som kan inträffa strax efter operationen. För det första finns det en risk för cementläckage, vilket kan orsaka lokala skador på den om-givande vävnaden, eller i sällsynta fall blodproppar. Dessutom tror man att användningen av bencementet kan underlätta uppkomsten av ytterligare frakturer i kotor intill den behandlade kotan, på grund av den höga styvheten hos cementet jämfört med den kringliggande ben-vävnaden, vilket kan ge en lokal förändring i hur lasterna i ryggen fördelas. De bencement som används idag har generellt en 2 till 20 gånger högre styvhet än den omgivande porösa benvävnaden, med större skillnader för benskört än för friskt ben.

Huvudsyftet med den här avhandlingen var därför att utveckla ett bencement med lägre styvhet, eftersom användningen av ett material

76

vars mekaniska egenskaper liknar de hos den porösa benvävnaden i kotorna skulle kunna bidra till att förebygga uppkomsten av nya frak-turer. Genom att tillsätta olika föreningar till kommersiellt tillgängliga akryliska bencement kan styvheten anpassas. Den första studien undersökte hur några funktionella egenskaper hos vanliga PMMA-cement påverkades av olika tillsatser i varierande mängder. Det visade sig att tillsatsen av fettsyror och triglyceridoljor förändrade hanteringsegenskaperna hos PMMA-cementen och orsa-kade en minskning av styvheten och hållfastheten, vilket var i linje med det övergripande målet. I synnerhet linolsyra (LA) verkade ha en större effekt på de mekaniska egenskaperna under tryck jämfört med de andra föreningarna, samtidigt som den inte gav någon negativ ef-fekt på hanteringsegenskaperna. LA-modifierade cement var därför i fortsatt fokus i denna avhandling, detta val stöddes också av tidigare lovande studier. Ytterligare funktionella egenskaper, såsom böjegenskaper, ut-dragskraften för skruvar i cementet och hanteringsegenskaper, utvär-derades i artikel II. På samma sätt som den tidigare påvisade påverkan på egenskaperna under tryck, uppvisade de LA-modifierade cementen också en lägre böjstyvhet och böjhållfasthet. Utdragskraften för skru-var var lägre med det modifierade cementet jämfört med det omodifie-rade, men värdena var liknande de som uppmätts för andra cementty-per såsom kalciumfosfat. Dessutom var det modifierade cementet möjligt att bearbeta under en längre tid jämfört med det omodifierade cementet, vilket innebär att läkaren har mer tid på sig att injicera ce-mentet in i den skadade kotan. Det experimentella cementet uppvisade även andra lämpliga hanteringsegenskaper, såsom liknande härdnings-tid och lägre polymerisationstemperatur (minskad risk för skador på kringliggande vävnader). Framställningen av en helt steril formulering genom autoklavering av LA hade ingen signifikant effekt på de funkt-ionella egenskaperna hos cementet. Med tanke på den avsedda kliniska applikationen så förväntas cemen-tet kunna motstå upprepad belastning in vivo. Av detta skäl utvärdera-

77

des cementens långsiktiga mekaniska prestanda i luft och i en saltlös-ning (37 °C), den senare för att bättre efterlikna in vivo-miljön. I arti-kel III uppvisade ett kommersiellt tillgängligt standardcement högre styvhet och hållfasthet än ett kommersiellt tillgängligt lågmodulce-ment. Denna skillnad härrördes till porositeten i lågmodulformule-ringen, vilken ökade över tid då cementen förvarades i saltlösning. På liknande sätt visade standardcementet en högre utmattningsgräns (47 MPa) än lågmodulcementet (31 MPa) i luft vid rumstemperatur. Båda utmattningsgränserna var betydligt högre än de belastningar som har uppmätts in vivo under dagliga fysiska aktiviteter (1-3 MPa). I artikel IV undersöktes utmattningsegenskaperna hos det experimen-tella cement som utvecklades i denna avhandling, i saltlösning vid 37 °C. På samma sätt som för tryckhållfastheten så var utmattningsgrän-sen för LA-modifierat cement väsentligt lägre än för det vanliga ce-mentet och minskade med nästan 90% (från 45 MPa till 4.7 MPa). Denna utmattningsgräns var i samma storleksordning som de spän-ningar som har uppmätts i ryggen in vivo, dock något högre. De modi-fierade cementen frigjorde en högre mängd monomer under härdning jämfört med kontrollcementet, vilket skulle kunna orsaka lokal kemisk skada på omkringliggande benvävnad. Det kommersiellt tillgängliga lågmodulcementet frigjorde dock liknande mängder, och en del andra cement i litteraturen ännu mer än detta. I in vivo-studien utförd i arti-kel V kunde dessutom ingen lokal celltoxicitet upptäckas. De modifie-rade och icke-modifierade cementen uppvisade liknande inflammato-riska reaktioner med bildandet av ett tunt fibröst membran vid gräns-snittet mellan benvävnad och bencement. Detta är ett typiskt svar på PMMA-cement i vävnader och visar att tillsatsen av små mängder LA till PMMA-bencement inte orsakar några ytterligare cytotoxiska ef-fekter på vävnaden runt implantationsställena. Sammanfattningsvis verkar LA-modifierade bencement vara ett lo-vande alternativ till de kommersiellt tillgängliga PMMA-cementen, för både behandling av kotkompressioner orsakade av benskörhet och som ett komplement till metallimplantat vid frakturfixering i bensköra patienter.

78

Acknowledgements

I would like to thank all the people who helped me during these four years and made this PhD experience such a positive and inspiring journey. In particular, I would like to thank the women I have been working with during these past years: my supervisors, Cecilia and Caroline, but also Malin, Gry and Marjam (even if we did not work together ☺). Thank you for being these strong and inspiring women, in all kind of ways! Cecilia, thank you for your unlimited patience, enthusiasm and ener-gy. Thank you for pushing me to levels I though I would never reach and for showing me what I am capable of. Caroline, thank you for your unlimited patience as well, and for al-ways being positive even when I doubted myself. Thank you for al-ways helping me with the fatigue measurements and for your kind-ness. To Håkan, thank you for giving me another perspective over Bio-materials and for the inspiring conversations during MIM Group meetings. Thank you for sharing your valuable thoughts and ideas. To Jöns, thank you for always taking time to answer to my questions and for the help with the papers. Gry, thank you for having so much energy and for always being in good mood. It warmed me when I needed it. To Malin, thank you for the support and the good advice during the in vivo study. Thank you for always being kind with me. Marjam, I would like to thank you for the nice conversations, great advice about life and for always bringing me positive vibes. Wei, thank you for your kindness and for always showing an interest when you meet me. JunJun, ☺ thank you for being, not only my officemate, but also my friend. Thank you for the absolute support and for always believing in

79

me, even when I did not myself. Thank you for sharing the chinese food and making me discover your culture through the stories of fa-mous Queens and Kings of China. To Shiuli & Michael (alias S&M), besides being true friends, thank you for your unconditional support. Michael, thank you for letting me be a part of your project and for always helping me. Thank you for your great sense of humor and for teaching me “never give up, never surrender” and (my favorite) “One shot, One kill”. Shiuli, thank you for being such a good person and for always waking up the romantic (and pathetic) side of my person ☺. Thank you for the good times with your family, in particular Ma. Special thanks to David, the Hero of the MIM lab ☺ Thank you for always helping me with all the smallest issues I could have. You al-ways have THE solutions, and I learnt a lot by your side. Ale, thank you for helping me with planning the dissertation party and for always answering to the questions related to polymers. Thank you for reminding me that “I am Latina”. Muchissimas gracias ☺! Thank you to Amina and Lee for always showing up at my office and asking if I was fine, (“you can do it!”). Thank you for your support ☺! To Le, thank you for the nice badminton games, and for sharing ideas for the presentations. I appreciated seeing you during the weekends, in the lab ☺. To Luimar, thank you for the nice talks and for being the best neighbor ever. To Susan, thank you for your enthusiasm and your support. Thank you for helping me with the English writing. ☺ To Ingrid A., thank you for showing your support and your kindness every time we meet. I really appreciated working with you as a master student. Anders P., thank you for the assistance with matlab, even when you did not have the time. Till Torbjörn, Peter, Amina, Camilla, Susanne, Charlotte, tack för att ni lärt mig svenska. Thank you for your friendship and for the nice Svensk Fika. Det var jättebra ! ☺ Thank you to Jonatan for your friendship and for always helping me when I had computer issues (I will request your help, no matter where I am). Thank you to Jocke (workshop), for always prioritizing my molds during the last year. Thank you to Janne, for the nice conversa-

80

tions about hunting and your help in all occasions. To Ingrid R., Sara R., Maria S., thank you for always helping me with the administrative procedures. To Ishtiaq, thank you for the company during the late night in the lab. To Céline (What a nice name!!!), thank you for com-ing to see me in June, for the defense. To all old and new members of the MIM group (Oscar, Dan, Julien, etc) and our division, thank you for your kindness. To Maria Pau, thank you for arranging the internship in Sweden 5 years ago. Thank you for introducing me to the world of biomaterials and for being my first supervisor. Thank you to Montserrat for your legendary kindness and always sharing your (unlimited) knowledge. Thank you for your warmth and empathy. To BIBITE’s group, thank you for the beautiful years in Barcelona. To my dear friends, outside the lab: Thank you to Marina, the two princesses & Carl for the nice family time and for the unconditional support ☺. Thank you Marie for the nice time at the gym, the numerous laughs and for being such an inspiring woman. Thank you, Marie & Pelle, for being some of my closest friends in Sweden. Special thanks to Yeelen, Liz and Larisa for all the crazy times togeth-er and for making this life an adventure ☺. Special thanks to Xi for the beautiful painting of the spine and the en-couragements. To Dibya & Camilla, thank you for the nice talks and all the nice moments together. Special thanks to Anna & Victoria, for always putting a smile on my face, no matter what. Thank you for the support, the help with the cover of the thesis and for the “I just came/call to sayyyyyy, I love youuuuuuuu”. Special Thanks to Igor & Kai, for everything!!! ☺ I am glad I got to know you and we became close enough friends to share plenty of times together. I love you guys and I will truly miss you! Special thanks for Sara, for her legendary kindness and friendship. Thank you for the late diners at your place and for always having the

81

warmest words. I truly appreciate the times we spend together. I hope there will be plenty more. Un spécial merci à ma chabinette Laétitia pour les jolis dessins de vertebroplasty, kyphoplasty et de vertèbre brisée. Je te souhaite beau-coup de courage ! Chrys, merci d’être venu me voir, j’apprécie le soutien que tu me portes et j’espère qu’on se verra plus souvent à l’avenir. Mamie, merci pour avoir entamer ce grand voyage depuis la Guyane, dans le seul but de venir m’écouter. J’apprécie le geste et j’espère que tu seras fière ! Jean-Gilles & Camille, il n’y a pas de mots pour décrire l’amour que je vous porte. Merci d’être mon frère et ma sœur chéris. Maman & Papa, merci pour votre inconditionnel soutien, amour, et tous les sacrifices que vous avez fait pour me permettre d’aller jusqu’au bout. Je sais que je ne le dis pas assez souvent: je vous aime plus que tout ! Tusen tack till alla mina kompisar!

82

References

(1) Ballane, G.; Cauley, J. A.; Luckey, M. M.; Fuleihan, G. E. Worldwide prevalence and incidence of osteoporotic vertebral fractures. Osteoporos. Int. 2017, 28 (5), 1531–1542 DOI: 10.1007/s00198-017-3909-3.

(2) Svedbom, A.; Hernlund, E.; Ivergård, M.; Compston, J.; Cooper, C.; Stenmark, J.; McCloskey, E. V.; Jönsson, B.; Kanis, J. A. Osteoporosis in the European Union: A compendium of country-specific reports. Arch. Osteoporos. 2013, 8 (1–2) DOI: 10.1007/s11657-013-0137-0.

(3) Rawls, A.; Fisher, R. E. Development and Functional Anatomy of the Spine. In The Genetics and Development of Scoliosis; 2010; pp 21–46.

(4) Adams, M. A. Biomechanics of back pain. Acupunct. Med. 2004, 22 (4), 178–188 DOI: 10.1136/aim.22.4.178.

(5) Ferguson, S. J. Biomechanics of the aging spine. 2003, 12, 97–103 DOI: 10.1007/s00586-003-0621-0.

(6) Adams, M. A.; Dolan, P. Spine biomechanics. J. Biomech. 2005, 38 (10), 1972–1983 DOI: 10.1016/j.jbiomech.2005.03.028.

(7) Myers, E. R.; Wilson, S. E. Biomechanics of Osteoporosis and Vertebral Fracture. Spine. 1997, p 25S–31S.

(8) Izzo, R.; Guarnieri, G.; Guglielmi, G.; Muto, M. Biomechanics of the spine. Part I: Spinal stability. Eur. J. Radiol. 2013, 82 (1), 118–126 DOI: 10.1016/j.ejrad.2012.07.024.

(9) Morgan, E. F.; Bayraktar, H. H.; Keaveny, T. M. Trabecular bone modulus-density relationships depend on anatomic site. J. Biomech. 2003, 36 (7), 897–904 DOI: 10.1016/S0021-9290(03)00071-X.

(10) Nazarian, A.; Von Stechow, D.; Zurakowski, D.; Müller, R.; Snyder, B. D. Bone volume fraction explains the variation in strength and stiffness of cancellous bone affected by metastatic cancer and osteoporosis. Calcif. Tissue Int. 2008, 83 (6), 368–379 DOI: 10.1007/s00223-008-9174-x.

(11) Wilke, H.; Neef, P.; Caimi, M.; Hoogland, T.; Claes, L. New in vivo measurements of pressure in the intervertebral disc in daily life. Spine (Phila. Pa. 1976). 1999, 8 (15), 755–762.

(12) Nachemson, A. L.; Elfström, G. Intravital Dynamic Pressure Measurements in Lumbar discs : A study of common movements, maneuvers and exercises. Scand. J. Rehabil. Med. Suppl. 1970, 1, 1–40.

(13) Burr, D. B.; Forwood, M. R.; Fyhrie, D. P.; Martin, R. B.; Schaffler, M. B.; Turner, C. H. Bone Microdamage and Skeletal Fragility in Osteoporotic and Stress Fractures. J. Bone Miner. Res. 1997, 12 (1), 6–15 DOI: 10.1359/jbmr.1997.12.1.6.

(14) Bowman, S. M.; Guo, X. E.; Cheng, D. W.; Keaveny, T. M.; Gibson, L. J.; Hayes, W. C.; MacMahon, T. A. Creep Contributes to the Fatigue Behavior of Bovine Trabecular bone. J. Biomech. Eng. 1998, 120 (5), 647–654 DOI: 10.1115/1.2834757.

83

(15) Haddock, S. M.; Yeh, O. C.; Mummaneni, P. V; Rosenberg, W. S.; Keaveny, T. M. Similarity in the fatigue behavior of trabecular bone across site and species. J. Biomech. 2004, 37 (2), 181–187 DOI: 10.1016/S0021-9290(03)00245-8.

(16) Michel, M. C.; Guo, X.-D. E.; Gibson, L. J.; McMahon, T. A.; Hayes, W. C. Compressive Fatigue Behavior of Bovine Trabecular Bone. J. Biomech. 1993, 26 (4), 453–463 DOI: 10.1016/0021-9290(93)90009-4.

(17) Verborgt, O.; Gibson, G. J.; Schaffler, M. B. Loss of Osteocyte Integrity in Association with Microdamage and Bone Remodeling After Fatigue In Vivo. J. Bone Miner. Res. 2000, 15 (1), 60–67 DOI: 10.1359/jbmr.2000.15.1.60.

(18) Watts, N. B. Osteoporotic vertebral fractures. Neurosurg. Focus 2001, 10 (4), 1–3.

(19) Schwartz, E. N.; Steinberg, D. Detection of Vertebral Fractures. Curr. Osteoporos. Rep. 2005, 3 (4), 126–135.

(20) Jalava, T.; Sarna, S.; Pylkka, L.; Selby, P.; Davies, M.; Adams, J.; Francis, R. M.; Robinson, J.; Mccloskey, E. Association between vertebral fracture and increased mortality in osteoporotic patients. J. Bone Miner. Res. 2003, 18 (7), 1254–1260.

(21) Longo, U. G.; Loppini, M.; Denaro, L.; Maffulli, N. Osteoporotic vertebral fractures : current concepts of conservative care. Br. Med. Bull. 2012, 102, 171–189 DOI: 10.1093/bmb/ldr048.

(22) Hernlund, E.; Svedbom, A.; Ivergård, M.; Compston, J.; Cooper, C.; Stenmark, J.; McCloskey, E. V.; Jönsson, B.; Kanis, J. A. Osteoporosis in the European Union: Medical management, epidemiology and economic burden. Arch. Osteoporos. 2013, 8 (1–2) DOI: 10.1007/s11657-013-0136-1.

(23) Etzioni, D. A.; Liu, J. H.; Maggard, M. A.; Ko, C. Y. The Aging Population and Its Impact on the Surgery Workforce. Ann. Surg. 2003, 238 (2), 170–177 DOI: 10.1097/00000658-200308000-00003.

(24) Bianchi, G.; Anderson, K. C. Understanding Biology to Tackle the Disease: Multiple Myeloma From Bench to Bedside, and Back. Ca Cancer J Clin 2014, 64 (6), 423–444 DOI: 10.3322/caac.21252.

(25) Mahindra, A.; Hideshima, T.; Anderson, K. C. Multiple myeloma: Biology of the disease. Blood Rev. 2010, 24 (SUPPL. 1), S5–S11 DOI: 10.1016/S0268-960X(10)70003-5.

(26) Vinay, S.; Khan, S.; Braybrooke, J. Lumbar vertebral haemangioma causing pathological fracture, epidural haemorrhage, and cord compression: a case report and review of literature. J. Spinal Cord Med. 2011, 34 (3), 335–339 DOI: 10.1179/2045772311Y.0000000004.

(27) Alfawareh, M.; Alotaibi, T.; Labeeb, A.; Audat, Z. A symptomatic case of thoracic vertebral hemangioma causing lower limb spastic paresis. Am. J. Case Rep. 2016, 17, 805–809 DOI: 10.12659/AJCR.898562.

(28) Kanis, J.; McCloskey, E.; Powles, T.; Paterson, A.; Ashley, S.; Spector, T. A high incidence of vertebral fracture in women with breast cancer. Br. J. Cancer 1999, 79 (7–8), 1179–1181 DOI: 10.1038/sj.bjc.6690188.

(29) Bartels, R. H. M. A.; van der Linden, Y. M.; van der Graaf, W. T. A. Spinal Extradural Metastasis: Review of Current Treatment Options. CA. Cancer J. Clin. 2008, 58 (4), 245–259 DOI: 10.3322/CA.2007.0016.

(30) Hipp, J. A.; Rosenberg, A. E.; Hayes, W. C. Mechanical properties of trabecular bone within and adjacent to osseous metastases. J. Bone Miner. Res. 1992, 7 (10), 1165–1171 DOI: 10.1002/jbmr.5650071008.

84

(31) Wong, C. C.; McGirt, M. J. Vertebral compression fractures: A review of current management and multimodal therapy. J. Multidiscip. Healthc. 2013, 6, 205–214 DOI: 10.2147/JMDH.S31659.

(32) Klazen, C. A. H.; Lohle, P. N. M.; De Vries, J.; Jansen, F. H.; Tielbeek, A. V.; Blonk, M. C.; Venmans, A.; Van Rooij, W. J. J.; Schoemaker, M. C.; Juttmann, J. R.; et al. Vertebroplasty versus conservative treatment in acute osteoporotic vertebral compression fractures (Vertos II): An open-label randomised trial. Lancet 2010, 376 (9746), 1085–1092 DOI: 10.1016/S0140-6736(10)60954-3.

(33) Miller, J. C. Augmentation Therapy for Vertebral Compression Fractures. Radiol. Rounds 2009, 7 (7), 1–4.

(34) Tohmeh, A. G.; Mathis, J. M.; Fenton, D. C.; Levine, A. M.; Belkoff, S. M. Biomechanical efficacy of Unipedicular vs bipedicular vertebroplasty for the management of osteoporotic compression fractures. Spine (Phila. Pa. 1976). 1999, 24 (17), 1772–1776.

(35) P. Galibert, H. Deramond, P. R. and D. L. G. Note préliminaire sur le traitement des angiomes vertébraux par vertébroplastie acrylique percutanée. Neurochirurgie. 1987, 33 (2), 166–168.

(36) Garfin, S. R.; Yuan, H. A.; Reiley, M. A. New technologies in spine: Kyphoplasty and vertebroplasty for the treatment of painful osteoporotic compression fractures. Spine (Phila. Pa. 1976). 2001, 26 (14), 1511–1515 DOI: 10.1097/00007632-200107150-00002.

(37) Lieberman, I. H.; Dudeney, S.; Reinhardt, M. K.; Bell, G. Initial outcome and efficacy of kyphoplasty in the treatment of painful osteoporotic vertebral compression fractures. Spine (Phila. Pa. 1976). 2001, 26 (14), 1631–1638 DOI: 10.1097/00007632-200107150-00026.

(38) Mukherjee, S.; Lee, Y. P. Current Concepts in the Management of Vertebral Compression Fractures. Oper. Tech. Orthop. 2011, 21 (3), 251–260 DOI: 10.1053/j.oto.2011.06.001.

(39) Kallmes, D. F.; Comstock, B. A.; Heagerty, P. J.; Turner, J. A.; Wilson, D. J.; Diamond, T. H.; Edwards, R.; Gray, L. A.; Stout, L.; Owen, S.; et al. A Randomized Trial of Vertebroplasty for Osteoporotic Spinal Fractures. N. Engl. J. Med. 2009, 361 (6), 569–579 DOI: 10.1056/NEJMoa0900563.

(40) R. Buchbinder, R.H. Osborne, B. M. A Randomized Trial of Vertebroplasty for painful Osteoporotic Vertebral Fractures. N. Engl. J. Med. 2009, 362 (6), 557–568.

(41) Papanastassiou, I. D.; Filis, A.; Gerochristou, M. A.; Vrionis, F. D. Controversial issues in kyphoplasty and vertebroplasty in osteoporotic vertebral fractures. Biomed Res. Int. 2014, 2014 (2) DOI: 10.1155/2014/934206.

(42) Barr, J. D.; Barr, M. S.; Lemley, T. J.; Mccann, R. M.; Cv, R. T. R. Percutaneous Vertebroplasty for Pain Relief and Spinal Stabilization. Spine (Phila. Pa. 1976). 2000, 25 (8), 923–928.

(43) Anselmetti, G. C.; Corrao, G.; Monica, P. Della; Tartaglia, V.; Manca, A.; Eminefendic, H.; Russo, F.; Tosetti, I.; Regge, D. Pain Relief Following Percutaneous Vertebroplasty : Results of a Series of 283 Consecutive Patients Treated in a Single. Cardiovasc. Intervent. Radiol. 2007, 30 (January), 441–447 DOI: 10.1007/s00270-006-0146-0.

(44) Grados, F.; Depriester, C.; Cayrolle, G.; Hardy, N.; Deramond, H.; Fardellone, P. Long-term observations of vertebral oeteoporotic fractures treated by percutaneous vertebroplasty. Rheumatology 2000, 39 (12), 1410–1414 DOI: 10.1093/rheumatology/39.12.1410.

85

(45) Mathis, J. M.; Barr, J. D.; Belkoff, S. M.; Barr, M. S.; Jensen, M. E. Percutaneous Vertebroplasty : A Developing Standard of Care for Vertebral Compression Fractures. Am. J. Neuroradiol. 2001, 22 (2), 373–381.

(46) Belkoff, S. M.; Mathis, J. M.; Jasper, L. E.; Deramond, H. The biomechanics of vertebroplasty: The effect of cement volume on mechanical behavior. Spine (Phila. Pa. 1976). 2001, 26 (14), 1537–1541 DOI: 10.1097/00007632-200107150-00007.

(47) Dash, A.; Brinster, D. R. Open heart surgery for removal of polymethylmethacrylate after percutaneous vertebroplasty. Ann. Thorac. Surg. 2011, 91 (1), 276–278 DOI: 10.1016/j.athoracsur.2010.06.106.

(48) Padovani, B.; Kasriel, O.; Brunner, P.; Peretti-viton, P. Pulmonary Embolism Caused by Acrylic Cement : A Rare Complication of Percutaneous Vertebroplasty. Am. J. Neuroradiol. 1999, 20 (3), 375–377.

(49) Scroop, R.; Eskridge, J.; Britz, G. W. Paradoxical cerebral arterial embolization of cement during intraoperative vertebroplasty: Case report. Am. J. Neuroradiol. 2002, 23 (5), 868–870.

(50) Uppin, A. A.; Hirsch, J. A.; Centenera, L. V; Pfiefer, B. A.; Pazianos, A. G. Occurrence of New Vertebral Body Fracture after Percutaneous Vertebroplasty in Patients with Osteoporosis. Radiology 2003, 226 (1), 6–8 DOI: 10.1148/radiol.2261011911.

(51) A.T Trout, D. F. K. and T. J. K. New fractures after vertebroplasty: adjacent fractures occur significantly sooner. Am. J. Neuroradiol. 2006, 27 (1), 217–223.

(52) Sun, Y.; Mu, M.; Teng, H.; Yuan, W.; Luo, C. Risk of post-vertebroplasty fracture in adjacent vertebral bodies appears correlated with the morphologic extent of bone cement. J. Chinese Med. Assoc. 2011, 74 (8), 357–362 DOI: 10.1016/j.jcma.2011.06.008.

(53) Nieuwenhuijse, M. J.; Putter, H.; van Erkel, A. R.; Dijkstra, P. D. S. New Vertebral Fractures after Percutaneous Vertebroplasty for Painful Osteoporotic Vertebral Compression Fractures: A Clustered Analysis and the Relevance of Intradiskal Cement Leakage. Radiology 2013, 266 (3), 862–870 DOI: 10.1148/radiol.12120751.

(54) Li, Y.-A.; Lin, C.-L.; Chang, M.-C.; Liu, C.-L.; Chen, T.-H.; Lai, S.-C. Subsequent Vertebral Fracture After Vertebroplasty. Spine (Phila. Pa. 1976). 2012, 37 (3), 179–183 DOI: 10.1097/BRS.0b013e3181f72b05.

(55) Tsoumakidou, G.; Too, C. W.; Koch, G.; Caudrelier, J.; Cazzato, R. L.; Garnon, J.; Gangi, A. CIRSE Guidelines on Percutaneous Vertebral Augmentation. Cardiovasc. Intervent. Radiol. 2017, 40 (3), 331–342 DOI: 10.1007/s00270-017-1574-8.

(56) Filippiadis, D. K.; Marcia, S.; Masala, S.; Deschamps, F.; Kelekis, A. Percutaneous Vertebroplasty and Kyphoplasty: Current Status, New Developments and Old Controversies. Cardiovasc. Intervent. Radiol. 2017, 40 (12), 1815–1823 DOI: 10.1007/s00270-017-1779-x.

(57) Kühn, K. Bone cements: Up-to-Date Comparison of Physical and Chemical Properties of Commercial Materials. In Springer; Berlin, 2000.

(58) Judet, J.; Judet, R. The use of an artificial femoral head for arthroplasty of the hip joint. J. Bone Joint Surg. Br. 1950, 32–B (2), 166–173.

(59) Haboush, E. J. A new operation for arthroplasty of the hip based on biomechanics, photoelasticity, fast setting dental acrylic, and other considerations. Bull. Hosp. Jt. Dis. 1953, 14 (2), 242–277.

(60) Charnley, J. The Classic: The bonding of Protheses to bone by Cement. J. Bone Jt. Surg. 1964, 46B, 518–529 DOI: 10.1007/s11999-010-1547-6.

86

(61) Lieberman, I. H.; Togawa, D.; Kayanja, M. M. Vertebroplasty and kyphoplasty : filler materials. Spine J. 2005, 5, 305S–316S DOI: 10.1016/j.spinee.2005.02.020.

(62) Lewis, G. Properties of Acrylic Bone Cement : State of the Art Review. 1997, No. May 1996.

(63) Lewis, G. Fatigue Testing and Performance of Acrylic Bone-Cement Materials : State-of-the-Art Review. J. Biomed. Mater. Res. - Part B 2003, 66B (1), 457–486.

(64) Iso 5833. Implants for surgery — Acrylic resin cements. 2002. (65) Lee, C. Properties of Bone Cement : The Mechanical Properties of PMMA

Bone Cement. Well-Cemented Total Hip Arthroplast. 2005, 60–66 DOI: 10.1007/3-540-28924-0_6.

(66) Ajaxon, I.; Persson, C. Compressive fatigue properties of a commercially available acrylic bone cement for vertebroplasty. Biomech. Model. Mechanobiol. 2014, 13 (6), 1199–1207 DOI: 10.1007/s10237-014-0566-8.

(67) Kurtz, S. M.; Villarraga, M. L.; Zhao, K.; Edidin, A. A. Static and fatigue mechanical behavior of bone cement with elevated barium sulfate content for treatment of vertebral compression fractures. Biomaterials 2005, 26 (17), 3699–3712 DOI: 10.1016/j.biomaterials.2004.09.055.

(68) Ishihara, S.; McEvily, A. J.; Goshima, T.; Kanekasu, K.; Nara, T. On fatigue lifetimes and fatigue crack growth behavior of bone cement. J Mater Sci Mater Med 2000, 11 (10), 661–666 DOI: 274378 [pii].

(69) ASTM F2118-10 (Reapproved 2009). Standard Test Method for Constant Amplitude of Force Controlled Fatigue Testing of. 2009, 1–18.

(70) Boelen, E. J. H.; Lewis, G.; Xu, J.; Slots, T.; Koole, L. H.; Van Hooy-Corstjens, C. S. J. Evaluation of a highly-radiopaque iodine-containing acrylic bone cement for use in augmentation of vertebral compression fractures. J. Biomed. Mater. Res. - Part A 2008, 86 (1), 76–88 DOI: 10.1002/jbm.a.31601.

(71) Lewis, G.; Koole, L. H.; Hooy-corstjens, C. S. J. Van. Influence of Powder-to-Liquid Monomer Ratio on Properties of an Injectable Iodine-Containing Acrylic Bone Cement for Vertebroplasty and Balloon Kyphoplasty. 2009, 6–10 DOI: 10.1002/jbm.b.31427.

(72) Öhman Mägi, C.; Holub, O.; Hall, R. M.; Persson, C. Density and mechanical properties of vertebral trabecular bone - a review. Manuscr. Draft.

(73) Harper, E. J.; Bonfield, W. Tensile characteristics of ten commercial acrylic bone cements. J. Biomed. Mater. Res. 2000, 53 (5), 605–616 DOI: 10.1002/1097-4636(200009)53:5<605::AID-JBM22>3.0.CO;2-5.

(74) Polikeit, A.; Nolte, L. P.; Ferguson, S. J. The effect of cement augmentation on the load transfer in an osteoporotic functional spinal unit: finite-element analysis. Spine (Phila. Pa. 1976). 2003, 28 (10), 991–996 DOI: 10.1097/01.BRS.0000061987.71624.17.

(75) Gutiérrez-Mejía, A.; Herrera-Kao, W.; Duarte-Aranda, S.; Loría-Bastarrachea, M. I.; Canché-Escamilla, G.; Moscoso-Sánchez, F. J.; Cauich-Rodríguez, J. V.; Cervantes-Uc, J. M. Synthesis and characterization of core-shell nanoparticles and their influence on the mechanical behavior of acrylic bone cements. Mater. Sci. Eng. C 2013, 33 (3), 1737–1743 DOI: 10.1016/j.msec.2012.12.087.

87

(76) Hong-Jiang, J.; Xu, J.; Qiu, Z.-Y.; Ma, X.-L.; Zhang, Z.-Q.; Tan, X.-X.; Cui, Y.; Cui, F.-Z. Mechanical properties and cytocompatibility improvement of vertebroplasty PMMA bone cements by incorporating mineralized collagen. Materials (Basel). 2015, 8 (5), 2616–2634 DOI: 10.3390/ma8052616.

(77) Vallittu, P. K.; Puska, M. A.; Kokkari, A. K.; Timo, O. N. Mechanical properties of oligomer-modified acrylic bone cement. Biomaterials 2003, 24, 417–425.

(78) Boger, A.; Bohner, M.; Heini, P.; Verrier, S.; Schneider, E. Properties of an Injectable Low Modulus PMMA Bone Cement for Osteoporotic Bone. J. Biomed. Mater. Res. - Part B Appl. Biomater. 2008, 86B (2), 474–482 DOI: 10.1002/jbm.b.31044.

(79) Lam, W. M.; Pan, H. B.; Fong, M. K.; Cheung, W. S.; Wong, K. L.; Li, Z. Y.; Luk, K. D. K.; Chan, W. K.; Wong, C. T.; Yang, C.; et al. In Vitro characterization of low modulus linoleic acid coated strontium- substituted hydroxyapatite containing PMMA bone cement. J. Biomed. Mater. Res. - Part B Appl. Biomater. 2010, 96B (1), 76–83 DOI: 10.1002/jbm.b.31741.

(80) Holub, O.; López, A.; Borse, V.; Engqvist, H.; Kapur, N.; Hall, R. M.; Persson, C. Biomechanics of low-modulus and standard acrylic bone cements in simulated vertebroplasty : A human ex vivo study. J. Biomech. 2015, 48 (12), 3258–3266 DOI: 10.1016/j.jbiomech.2015.06.026.

(81) Kinzl, M.; Benneker, L. M.; Boger, A.; Zysset, P. K.; Pahr, D. H. The effect of standard and low-modulus cement augmentation on the stiffness, strength, and endplate pressure distribution in vertebroplasty. Eur. Spine J. 2012, 21 (5), 920–929 DOI: 10.1007/s00586-011-2119-5.

(82) Kim, J. M.; Shin, D. A.; Byun, D. H.; Kim, H. S.; Kim, S.; Kim, H. I. Effect of bone cement volume and stiffness on occurrences of adjacent vertebral fractures after vertebroplasty. J. Korean Neurosurg. Soc. 2012, 52 (5), 435–440 DOI: 10.3340/jkns.2012.52.5.435.

(83) Boger, A.; Bisig, A.; Bohner, M.; Heini, P.; Schneider, E. Variation of the mechanical properties of PMMA to suit osteoporotic cancellous bone. J. Biomater. Sci. Polym. Ed. 2008, 19 (9), 1125–1142 DOI: 10.1163/156856208785540154.

(84) Harper, E. J.; Behiri, J. C.; Bonfield, W. Flexural and fatigue properties of a bone cement based upon polyethylmethacrylate and hydroxyapatite. J. Mater. Sci. Mater. Med. 1995, 6 (12), 799–803 DOI: 10.1007/BF00134320.

(85) Weightman, B.; Freeman, M. a; Revell, P. a; Braden, M.; Albrektsson, B. E.; Carlson, L. V. The mechanical properties of cement and loosening of the femoral component of hip replacements. J. Bone Joint Surg. Br. 1987, 69 (4), 558–564.

(86) A. López, G. Mestres, M. Karlsson Ott, H. Engqvist, S. J. Ferguson, C. P. and B. H. Compressive mechanical properties and cytocompatibility of bone-compliant , linoleic acid-modified bone cement in a bovine model. J. Mech. Behav. Biomed. Mater. 2014, 32, 245–256 DOI: 10.1016/j.jmbbm.2014.01.002.

(87) Boger, A.; Wheeler, K.; Montali, A.; Gruskin, E. NMP-modified PMMA bone cement with adapted mechanical and hardening properties for the use in cancellous bone augmentation. J. Biomed. Mater. Res. - Part B Appl. Biomater. 2009, 90 (2), 760–766 DOI: 10.1002/jbm.b.31345.

(88) López, A.; Hoess, A.; Thersleff, T.; Ott, M.; Engqvist, H.; Persson, C. Low-modulus PMMA bone cement modified with castor oil. Biomed. Mater. Eng. 2011, 21 (5–6), 323–332 DOI: 10.3233/BME-2012-0679.

88

(89) Persson, C.; Lopez, A.; Fathali, H.; Hoess, A.; Rojas, R.; Ott, M. K.; Hilborn, J.; Engqvist, H. The effect of oligo(trimethylene carbonate) addition on the stiffness of acrylic bone cement. Biomatter 2016, 6 (1), 1–8 DOI: 10.1080/21592535.2015.1133394.

(90) Kolb, J. P.; Kueny, R. A.; Püschel, K.; Boger, A.; Rueger, J. M.; Morlock, M. M.; Huber, G.; Lehmann, W. Does the cement stiffness affect fatigue fracture strength of vertebrae after cement augmentation in osteoporotic patients? Eur. Spine J. 2013, 22 (7), 1650–1656 DOI: 10.1007/s00586-013-2809-2.

(91) Belkoff, S. M.; Mathis, J. M.; Erbe, E. M.; Fenton, D. C. Biomechanical evaluation of a new bone cement for use in vertebroplasty. Spine (Phila. Pa. 1976). 2000, 25 (9), 1061–1064 DOI: 10.1097/00007632-200005010-00004.

(92) Comparison, M.; Polymethylmethacrylate, O. F. Histological, radiological and mechanical comparison of polymethylmethacrylate (PMMA) and a bioactive bone cement in an ovine model. (Abstract) 1999, No. 610, 503.

(93) Belkoff, S. M.; Mathis, J. M.; Jasper, L. E.; Deramond, H. An ex vivo biomechanical evaluation of a hydroxyapatite cement for use with vertebroplasty. Spine (Phila. Pa. 1976). 2001, 26 (14), 1542–1546 DOI: 10.1097/00007632-200107150-00008.

(94) Heini, P. F.; Berlemann, U. Bone substitutes in vertebroplasty. Eur Spine J 2001, 10 (0), 205–213 DOI: 10.1007/s005860100308.

(95) He, Z.; Zhai, Q.; Hu, M.; Cao, C.; Wang, J.; Yang, H.; Li, B. Bone cements for percutaneous vertebroplasty and balloon kyphoplasty: Current status and future developments. J. Orthop. Transl. 2015, 3 (1), 1–11 DOI: 10.1016/j.jot.2014.11.002.

(96) Persson, C.; Robert, E.; Carlsson, E.; Robo, C.; Lopez, a.; Godoy-Gallardo, M.; Ginebra, M.-P.; Engqvist, H. The effect of unsaturated fatty acid and triglyceride oil addition on the mechanical and antibacterial properties of acrylic bone cements. J. Biomater. Appl. 2015, 0 (0), 1–11 DOI: 10.1177/0885328215581316.

(97) Güner, F. S.; Yagci, Y.; Erciyes, A. T. Polymers from triglyceride oils. Prog. Polym. Sci. 7 2006, 31 (7), 633–670 DOI: 10.1016/j.progpolymsci.2006.07.001.

(98) Mutlu, H.; Meier, M. A. R. Castor oil as a renewable resource for the chemical industry. Eur. J. Lipid Sci. Technol. 2010, 112 (1), 10–30 DOI: 10.1002/ejlt.200900138.

(99) Lazzari, M.; Chiantore, O. Drying and oxidative degradation of linseed oil. Polym. Degrad. Stab. 1999, 65 (2), 303–313 DOI: 10.1016/S0141-3910(99)00020-8.

(100) Sharma, V.; Kundu, P. P. Addition polymers from natural oils-A review. Prog. Polym. Sci. 2006, 31 (11), 983–1008 DOI: 10.1016/j.progpolymsci.2006.09.003.

(101) ISO 5833: Implants for surgery-Acrylic resin cements. 2002. (102) Lewis, G. Properties of Acrylic Bone Cement : State of the Art Review. J.

Biomed. Mater. Res. - Part A 1997, 38 (2), 155–182 DOI: 10.1002/(SICI)1097-4636(199722)38:2<155::AID-JBM10>3.0.CO;2-C.

(103) Tanner, K. E.; Wang, J.-S.; Kjellson, F.; Lidgren, L. Comparison of two methods of fatigue testing bone cement. Acta Biomater. 2010, 6 (3), 943–952 DOI: 10.1016/j.actbio.2009.09.009.

(104) Van Hooy-Corstjens, C. S. J.; Govaert, L. E.; Spoelstra, A. B.; Bulstra, S. K.; Wetzels, G. M. R.; Koole, L. H. Mechanical behaviour of a new acrylic radiopaque iodine-containing bone cement. Biomaterials 2004, 25 (13), 2657–2667 DOI: 10.1016/j.biomaterials.2003.09.038.

89

(105) Lewis, G.; Koole, L. H.; Van Hooy-Corstjens, C. S. J. Influence of powder-to-liquid monomer ratio on properties of an injectable iodine-containing acrylic bone cement for vertebroplasty and balloon kyphoplasty. J. Biomed. Mater. Res. - Part B Appl. Biomater. 2009, 91 (2), 537–544 DOI: 10.1002/jbm.b.31427.

(106) Lewis, G.; Janna, S. Effect of test frequency on the in vitro fatigue life of acrylic bone cement. Biomaterials 2003, 24 (23), 4315–4321 DOI: 10.1016/S0142-9612(03)00316-8.

(107) Köster, U.; Jaeger, R.; Bardts, M.; Wahnes, C.; Büchner, H.; Kühn, K. D.; Vogt, S. Creep and fatigue behavior of a novel 2-component paste-like formulation of acrylic bone cements. J. Mater. Sci. Mater. Med. 2013, 24 (6), 1395–1406 DOI: 10.1007/s10856-013-4909-2.

(108) Serbetci, K.; Korkusuz, F.; Hasirci, N. Thermal and mechanical properties of hydroxyapatite impregnated acrylic bone cements. Polym. Test. 2004, 23 (2), 145–155 DOI: 10.1016/S0142-9418(03)00073-4.

(109) Fritsch, E. W. Static and fatigue properties of two new low-viscosity PMMA bone cements improved by vacuum mixing. J. Biomed. Mater. Res. 1996, 31, 451–456 DOI: 10.1002/(SICI)1097-4636(199608)31:4&lt.

(110) Lewis, G.; Mladsi, S. Effect of sterilization method on properties of Palacos R acrylic bone cement. Biomaterials 1998, 19 (1–3), 117–124 DOI: 10.1016/S0142-9612(97)00165-8.

(111) W.R. Krause, L.W. Grimes, R. S. M. Fatigue testing of acrylic bone cements : Statistical concepts and proposed test methodology. J. Biomed. Mater. Res. - Part A 1988, 22 (A2), 179–190.

(112) L. Cristofolini, C. Minari, M. Vi. A methodology and criterion for acrylic bone cement fatigue tests. Fatigue Fract. Eng. Mater. Struct. 2000, 23 (11), 953–957.

(113) Baleani, M.; Persson, C.; Guandalini, L.; Piazza, E.; Erani, P.; Squarzoni, S.; Viceconti, M. The effect of gentamicin sulphate on the fracture properties of a manually mixed bone cement. Fatigue Fract. Eng. Mater. Struct. 2007, 30 (6), 479–488.

(114) Nakazawa, H.; Kodama, S. Statistical S-N testing method with 14 specimens: JSME standard Method for determination of S-N curves. Stat. Res. Fatigue Fract. Curr. Japanese Mater. Res. 1987, 2, 59–69.

(115) Dixon, W. J.; Mood, A. M. A Method for obtaining and analyzing sensitivity data. J. to Am. Assoc. 1948, 43 (241), 109–126.

(116) Bruce, R. D. An Up-and-Down for Acute Toxicity Testing. Fundam. Appl. Toxicol. 1985, 5, 151–157.

(117) Dixon, W. J. The Up-and-Down Method for Small Samples. J. Am. Stat. Assoc. 1965, 60 (312), 967–978.

(118) Brownlee, K. A.; Hodges, J. L.; Rosenblatt, J. and M. The Up-and-Down Method with Small Samples. J. Am. Stat. Assoc. 1953, 48 (262), 262–277.

(119) Bohner, M.; Baroud, G. Injectability of calcium phosphate pastes. Biomaterials 2005, 26 (13), 1553–1563 DOI: 10.1016/j.biomaterials.2004.05.010.

(120) Dahl, O. E.; Garvik, L. J.; Lyberg, T. Toxic effects of methylmethacrylate monomer on leukocytes and endothelial cells in vitro. Acta Orthop. 1994, 65 (2), 147–153 DOI: 10.3109/17453679408995423.

90

(121) Moreau, M. F.; Chappard, D.; Lesourd, M.; Monthéard, J. P.; Baslé, M. F. Free radicals and side products released during methylmethacrylate polymerization are cytotoxic for osteoblastic cells. J. Biomed. Mater. Res. 1998, 40 (1), 124–131 DOI: 10.1002/(SICI)1097-4636(199804)40:1<124::AID-JBM14>3.0.CO;2-O.

(122) Leggat, P. A. Toxicity of methyl methacrylate in dentistry. Int. Dent. J. 2003, 53, 126–131.

(123) ASTM F451-08. Standard Specification for Acrylic Bone Cement. 2008, 1–11.

(124) Sarzier, J. S.; Evans, A. J.; Cahill, D. W. Increased pedicle screw pullout strength with vertebroplasty augmentation in osteoporotic spines. J. Neurosurg. 2002, 96 (3 Suppl), 309–312 DOI: 10.3171/spi.2002.96.3.0309.

(125) Cook, S. D.; Salkeld, S. L.; Stanley, T.; Faciane, A.; Miller, S. D. Biomechanical study of pedicle screw fixation in severely osteoporotic bone. Spine J. 2004, 4 (4), 402–408 DOI: 10.1016/j.spinee.2003.11.010.

(126) F543-07, A. Standard Specification and Test Methods for Metallic Medical Bone Screws. 2007, 1–20 DOI: 10.1520/F0543-07E01.

(127) ISO 10993-6: Biological evaluation of medical devices (Part 6: Tests for local effects after implantation). 2007.

(128) Baroud, G.; Bohner, M.; Heini, P.; Steffen, T. Injection biomechanics of bone cements used in vertebroplasty. Biomed. Mater. Eng. 2004, 14 (4), 487–504.

(129) Krebs, J.; Ferguson, S. J.; Bohner, M.; Baroud, G.; Steffen, T.; Heini, P. F. Clinical Measurements of Cement Injection Pressure During Vertebroplasty. Spine (Phila. Pa. 1976). 2005, 30 (5), E118–E122 DOI: 00007632-200503010-00022 [pii].

(130) Baroud, G.; Vant, C.; Giannitsios, D.; Bohner, M.; Steffen, T. Effect of vertebral shell on injection pressure and intravertebral pressure in vertebroplasty. Spine (Phila. Pa. 1976). 2005, 30 (1), 68–74 DOI: 00007632-200501010-00013 [pii].

(131) Bohner, M.; Gasser, B.; Baroud, G.; Heini, P. Theoretical and experimental model to describe the injection of a polymethylmethacrylate cement into a porous structure. Biomaterials 2003, 24 (16), 2721–2730 DOI: 10.1016/S0142-9612(03)00086-3.

(132) Pujari-Palmer, M.; Robo, C.; Persson, C.; Procter, P.; Engqvist, H. Influence of cement compressive strength and porosity on augmentation performance in a model of orthopedic screw pull-out. J. Mech. Behav. Biomed. Mater. 2018, 77 (August 2017), 624–633 DOI: 10.1016/j.jmbbm.2017.10.016.

(133) Carlsson, E.; Mestres, G.; Treerattrakoon, K.; López, A.; Ott, M. K.; Larsson, S.; Persson, C. In Vitro and In Vivo Response to Low-Modulus PMMA-Based Bone Cement. Biomed Res. Int. 2015, 2015, 1–9 DOI: 10.1155/2015/594284.

(134) Guo, J.; Schork, F. J. Hybrid Miniemulsion Polymerization of Acrylate / Oil and Acrylate / Fatty Acid Systems. Macromol. React. Eng. 2008, 2 (3), 265–276 DOI: 10.1002/mren.200800004.

(135) Gurgel Adeodato Vieira, M.; Altenhofen Da Silva, M.; Oliveira Dos Santos, L.; Masumi Beppu, M. Natural-based plasticizers and biopolymer films : A review. Eur. Polym. J. 2011, 47 (3), 254–263 DOI: 10.1016/j.eurpolymj.2010.12.011.

(136) F. X. Budi Santosa, G. W. P. Tensile properties and Water absorption of Zein Sheets Plasticized with Oleic acid and Linoleic acid. J. Agric. Food Chem. 1999, 47 (5), 2070–2074 DOI: 10.1021/jf981154p.

91

(137) Daniels, P. H. A Brief Overview of Theories of PVC Plasticization and Methods Used to Evaluate PVC-Plasticizer Interaction. J. Vinyl Addit. Technol. 2009, 15 (4), 219–223 DOI: 10.1002/vnl.20211.

(138) Vallo, C. I.; Montemartini, P. E.; Cuadrado, T. R. Effect of Residual Monomer Content on Some Properties of a Poly ( methyl methacrylate ) -Based Bone Cement. J. Appl. Pol. Sci. 1998, No. 7600, 1367–1383.

(139) Milner, R. The Development of Theoretical Relationships between Some Handling Parameters (Setting Time and Setting Temperature), Composition (Relative Amounts of Initiator and Activator) and Ambient Temperature for Acrylic Bone Cement. J. Biomed. Mater. Res. - Part B 2004, 68B (2), 180–185 DOI: 10.1002/jbm.b.20019.

(140) Mjöberg, B.; Pettersson, H.; Rosenqvist, R.; Rydholm, A. Bone cement , thermal injury and the radiolucent zone. Acta Orthop. Scand. 1984, 55 (6), 597–600 DOI: 10.3109/17453678408992403.

(141) Leeson, M. C.; Lippitt, S. B. Thermal Aspects of the Use of Polymethacrylate in Large Methaphyseal Defects in Bone: A clinical review and Laboratory study. Clin. Orthop. Relat. Res. 1993, 295, 239–245 DOI: 10.1007/978-4-431-68072-7_42.

(142) Belkoff, S. M.; Molloy, S. Temperature Measurement During Polymerization of Polymethylmethacrylate Cement Used for Vertebroplasty. Spine (Phila. Pa. 1976). 2003, 28 (14), 1555–1559.

(143) Jefferiss, C. D.; Lee, A. J.; Ling, R. S. Thermal aspects of self-curing polymethylmethacrylate. J. Bone Joint Surg. Br. 1975, 57 (4), 511–518.

(144) Li, C.; Mason, J.; Yakimicki, D. Thermal characterization of PMMA-based bone cement curing. J. Mater. Sci. Mater. Med. 2004, 15 (1), 85–89 DOI: 10.1023/B:JMSM.0000010101.45352.d1.

(145) Deramond, H.; Wright, N. T.; Belkoff, S. M. Temperature Elevation Caused by Bone Cement Polymerization During Vertebroplasty. Bone 1999, 25 (2), 17S–21S DOI: 10.1016/S8756-3282(99)00127-1.

(146) San Millán Ruíz, D.; Burkhardt, K.; Jean, B.; Muster, M.; Martin, J. B.; Bouvier, J.; Fasel, J. H. D.; Rüfenacht, D. A.; Kurt, A. M. Pathology findings with acrylic implants. Bone 1999, 25 (SUPPL. 1), 0–5 DOI: 10.1016/S8756-3282(99)00140-4.

(147) Wendal, K.; Scheuermann, H.; Weitzel, E.; Rudigier, J. Pharmacokinetics of methylmethacrylate monomer during total hip replacement in man. Arch. Orthop. Trauma. Surg. 1988, 107 (5), 316–321 DOI: 10.1007/BF00451513.

(148) Baroud, G.; Steffen, T. A new cannula to ease cement injection during vertebroplasty. Eur. Spine J. 2005, 14 (5), 474–479 DOI: 10.1007/s00586-004-0822-1.

(149) Jensen, M. E.; Evans, A. J.; Mathis, J. M.; Kallmes, D. F.; Cloft, H. J.; Dion, J. E. Percutaneous polymethylmethacrylate vertebroplasty in the treatment of osteoporotic vertebral body compression fractures: Technical aspects. Am. J. Neuroradiol. 1997, 18 (10), 1897–1904.

(150) Martin, J. B.; Jean, B.; Sugiu, K.; San Millán Ruíz, D.; Piotin, M.; Murphy, K.; Rüfenacht, B.; Muster, M.; Rüfenacht, D. A. Vertebroplasty : Clinical Experience and Follow-up Results. Bone 1999, 25 (2), 11–15 DOI: 10.1016/S8756-3282(99)00126-X.

(151) Breusch, S.; Heisel, C.; Müller, J.; Borchers, T.; Mau, H. Influence of cement viscosity on cement interdigitation and venous fat content under in vivo conditions: A bilateral study of 13 sheep. Acta Orthop. Scand. 2002, 73 (4), 409–415 DOI: 10.1080/00016470216320.

92

(152) Kobayashi, N.; Togawa, D.; Fujishiro, T.; Powell, K. A.; Turner, A. S.; Seim, H. B.; Bauer, T. W. Histological and radiographic evaluation of polymethylmethacrylate with two different concentrations of barium sulfate in a sheep vertebroplasty model. J. Biomed. Mater. Res. - Part A 2005, 75 (1), 123–127 DOI: 10.1002/jbm.a.30388.

(153) M. A. R. Freeman; G. W. Bradley; P. A. Revell. Observations upon the interface between bone and polymethylmethacrylate cement. Bone Jt. J. 1982, 64 (4), 489–493.

(154) Wooley, P. H.; Morren, R.; Andary, J.; Sud, S.; Yang, S.; Markel, D.; Sieving, A.; Nasser, S. Inflammatory responses to orthopaedic biomaterials in the murine air pouch. Biomaterials 2002, 23 (2), 517–526 DOI: 10.1016/S0142-9612(01)00134-X.

(155) J. Charnley. The reaction of bone to self-curing acrylic cement: A long term Histological Study in Man. J. bone Jt. Surg. 1970, 52 (2), 340–353.

(156) Goodman, S. B.; Schatzker, J.; Sumner-Smith, G.; Fornasier, V. L.; Goften, N.; Hunt, C. The Effect of Polymethylmethacrylate on Bone : an Experimental Study. Arch. Orthop. Trauma. Surg. 1985, 104 (3), 150–154 DOI: 10.1007/BF00454691.

(157) Togawa, D.; Bauer, T. W.; Lieberman, I. H.; Takikawa, S. Histologic Evaluation of Human Vertebral Bodies after Vertebral Augmentation with Polymethyl Methacrylate. Spine (Phila. Pa. 1976). 2003, 28 (14), 1521–1527.

(158) Nottrott, M.; Mølster, A. O.; Moldestad, I. O.; Walsh, W. R.; Gjerdet, N. R. Performance of bone cements: Are current preclinical specifications adequate? Acta Orthop. 2008, 79 (6), 826–831 DOI: 10.1080/17453670810016920.

(159) Wilke, H. J.; Neef, P.; Caimi, M.; Hoogland, T.; Claes, L. E. New In Vivo Measurements of Pressures in the Intervertebral Disc in Daily Life. Spine (Phila. Pa. 1976). 1999, 24 (8), 755–762.

(160) Nachemson, A. L. Disc Pressure Measurements. Spine (Phila. Pa. 1976). 1981, 6 (1), 93–97.

(161) Takahara, K.; Kamimura, M.; Moriya, H.; Ashizawa, R.; Koike, T.; Hidai, Y.; Ikegami, S. Risk factors of adjacent vertebral collapse after percutaneous vertebroplasty for osteoporotic vertebral fracture in postmenopausal women. BMC Musculoskelet. Disord. 2016, 17 (12), 1–7 DOI: 10.1186/s12891-016-0887-0.

(162) Lo, Y.-P.; Chen, W.-J.; Chen, L.-H.; Lai, P.-L. New Vertebral Fracture After Vertebroplasty. J. Trauma an Acute Care Surg. 2008, 65 (6), 1439–1445 DOI: 10.1097/TA.0b013e318169cd0b.

(163) Bae, J. S.; Park, J. H.; Kim, K. J.; Kim, H. S.; Jang, I. T. Analysis of Risk Factors for Secondary New Vertebral Compression Fracture Following Percutaneous Vertebroplasty in Patients with Osteoporosis. World Neurosurg. 2017, 99, 387–394 DOI: 10.1016/j.wneu.2016.12.038.

(164) Stadelmann, V. A.; Bretton, E.; Terrier, A.; Procter, P.; Pioletti, D. P. Calcium phosphate cement augmentation of cancellous bone screws can compensate for the absence of cortical fixation. J. Biomech. 2010, 43 (15), 2869–2874 DOI: 10.1016/j.jbiomech.2010.07.025.

(165) Juvonen, T.; Nuutinen, J. P.; Koistinen, A. P.; Kröger, H.; Lappalainen, R. Biomechanical evaluation of bone screw fixation with a novel bone cement. Biomed. Eng. Online 2015, 14 (1), 3–12 DOI: 10.1186/s12938-015-0069-6.

Acta Universitatis UpsaliensisDigital Comprehensive Summaries of Uppsala Dissertationsfrom the Faculty of Science and Technology 1676

Editor: The Dean of the Faculty of Science and Technology

A doctoral dissertation from the Faculty of Science andTechnology, Uppsala University, is usually a summary of anumber of papers. A few copies of the complete dissertationare kept at major Swedish research libraries, while thesummary alone is distributed internationally throughthe series Digital Comprehensive Summaries of UppsalaDissertations from the Faculty of Science and Technology.(Prior to January, 2005, the series was published under thetitle “Comprehensive Summaries of Uppsala Dissertationsfrom the Faculty of Science and Technology”.)

Distribution: publications.uu.seurn:nbn:se:uu:diva-349028

ACTAUNIVERSITATIS

UPSALIENSISUPPSALA

2018