Compound meniscus
implant prototypes
Bench test performance of knitted casing to contain, fixate and
mechanically stabilize cell seeded gels
Maria Ydrefors Thesis for the Degree of Master in Science
with a major in Textile Engineering The Swedish School of Textiles
ii
ABSTRACT
Meniscal tears are the most common intra-articular injury of the knee joint. Due to the avascular zone with limited blood supply, treatment of the injury is a complex process. Today, research on the development of efficient treatments and meniscal replacements is of increasing interest. However, there are few alternatives of meniscal replacements available on the market and research has shown uncertain results in their ability to restore the natural biomechanics of the knee joint or prevent development of osteoarthritis. Furthermore there is no comparable method to evaluate tensile stresses caused by axial compressional load on a whole meniscus replacement. Therefore the possibility of knitted casing to contain, fixate and mechanically stabilize a cell seeded bioprinted gel and develop a methodology to characterize its compressional behaviour was analysed. By interlock knitting with segments of partial knit a 3D crescent-shaped biodegradable casing was produced mimicking the dimension of the medial meniscus. In the casing design, an Artelon® Flexband™ was incorporated functioning both as reinforcement at the peripheral rim and as fixation method. Moreover radial threads were added to the casing design by inclusion of weft inlays in the knitting pattern. In the non-destructive characterization of the compressional behaviour of the prototype, axial compressional forces of 10.82 N and 29.77 N were achieved. However the forces achieved were significantly lower if compared to the high force that is applied to the menisci in the knee joint. Furthermore a high influence of the coefficient of friction of the casing in the axial compressional force was concluded. Nevertheless refinements of the methodology are required to perform evaluation with comparable and reliable results.
Keywords: Knitting technology, meniscus replacement, hydrogel, Artelon®,
iii
POPULAR ABSTRACT
iv
ACKNOWLEDGEMENTS
First of all I would like to thank my supervisor Anders Persson for initiating the topic and for guiding me through this thesis. Thank you for patiently listening to all my questions and for all the input on both the theoretical and the practical work. I also want to thank my supervisor in knitting technology, Kristian Rödby, for all the hours you have spent on the development of the knitting patterns and the production of the knitted casing samples.
Special thanks to the researchers at Sahlgrenska Academy, Docent, Stina Simonsson, and Carl Lindahl PhD, for providing with bioprinted gels and for your support during this project. Without your research, I would have not had this thesis topic. Further, I would like to thank Prof Emer. Lars Peterson for your guidance in the design and the necessary requirements of a meniscal replacement. Additionally I would like to thank Aaron Smith, CEO of Artelon® for your generosity in providing the project with the necessary materials.
Finally, I would like to thank my family and friends for all your love and support. I could not have done this without you. A special thanks to Emanuel Ydrefors, for always supporting me in my studies and for your generous help in proof-reading and guidance in the thesis writing. Lastly I would like to thank my lovely classmates for all the encouragements and all the fun lunches and activities during these two years in the master programme.
v
TERMINOLOGY
Allograft= Cell, tissue or organ from donor
Anterior = Anatomical term for front, in this case the meniscus tip facing the front
of the body
Articular surface= The surface of a synovial joint where the ends of the
connecting bones meet. The articular surfaces are usually covered with articular cartilage.
Avascular = Tissue without blood vessels
Cell seeding = Insertion of cells into an extracellular matrix or host tissue, in this
case stem cells
Condyle = The round shaped prominence at the end of a bone, often in articulation
with a connecting bone.
ECM = Extracellular Matrix, a mesh-like substance in the extracellular space,
which provides structural properties and anchorage to the other tissues and cells.
Femur =Thigh bone Fibula = Calf bone
In vivo = In a living matter In vitro = In a glass ware
Lateral = To the side, the external side of the body
Medial = Inner side, the side closer to the centre of the body
Meniscectomy= To surgically remove the complete or partial of the meniscus. MRI= Magnet Resonance Imaging, A non-invasive diagnostic imaging technique to study soft tissues.
Osteoarthritis= A degenerative joint disease caused by degradation and attrition of
cartilage.
Patella= knee cap
Peripheral rim= Here, the external edge of the meniscus. Opposite the internal
circumference.
Posterior = Anatomical term for back, in this case the meniscus tip facing the
backside of the body
Synovial membrane = A membrane impermeable to a synovial fluid Suture= Surgical seam to join body tissue.
Tibia = Shin bone
vi
TABLE OF CONTENT
ABSTRACT ... ii
POPULAR ABSTRACT ... iii
ACKNOWLEDGEMENTS ... iv TERMINOLOGY ... v 1. INTRODUCTION ... 1 1.1 PROBLEM DESCRIPTION ... 3 1.2 AIM ... 4 1.3 REQUIREMENTS ... 5 1.4 RESEARCH QUESTIONS ... 5 1.5 LIMITATIONS ... 6 1.6 LITERATURE REVIEW ... 7
1.6.1 THE KNEE JOINT ... 7
1.6.2 THE MENISCUS ... 9
1.6.2.1 ANATOMY OF THE MENISCUS ... 9
1.6.2.2 BIOCHEMICAL COMPOSITION OF THE MENISCUS ... 10
1.6.2.3 BIOMECHANICAL PROPERTIES OF THE MENISCUS ... 10
1.6.3 MENISCAL INJURIES ... 12
1.6.4 CONSTRUCTION OF A MENISCUS REPLACEMENT ... 13
1.6.5 EVALUATION METHODS FOR BIOMECHANICAL PERFORMANCE OF MENISCUS REPLACEMENTS ... 14
1.6.6 BIOPRINTING AND HYDROGEL ... 15
1.6.7 ARTELON® ... 16
1.6.8 KNITTING TECHNOLOGY ... 16
2. EXPERIMENTAL SECTION ... 17
2.1 MATERIAL ... 17
2.1.1 CASING MATERIAL ... 17
2.1.2 BIOPRINTED GEL AND PROTOTYPES ... 18
2.2 FABRICATION METHOD ... 19
2.2.1 BIOPRINTED GEL PROTOTYPES ... 19
2.2.3 HEAT SETTING OF CASINGS ... 20
2.2.3.1 HEAT SETTING OF POLYESTER CASINGS ... 21
2.2.3.2 HEAT SETTING OF ARTELON CASINGS ... 21
2.3 CHARACTERIZATION METHODS ... 21
2.3.1 VALIDATION AND FRAME OF REFERENCE TESTING ... 21
vii
2.3.3 CALCULATION OF AXIAL COMPRESSIONAL FORCE ... 24
2.3.4 DETERMINATION OF COEFFICIENT OF FRICTION ... 25
2.3.5 LEAKAGE TEST ... 27
2.4 STATISTICAL METHODS ... 27
3. RESULT... 28
3.1 FABRICATION METHODS... 28
3.1.1 DESIGN AND KNITTING OF CASINGS ... 28
3.1.2 HEAT SETTING OF CASINGS ... 30
3.2 CHARACTERIZATION AND STATISTICAL ANALYSIS ... 31
3.2.1 EVALUATION OF COEFFICIENT OF FRICTION ... 31
3.2.2 EVALUATION OF AXIAL COMPRESSIONAL BEHAVIOUR ... 32
3.2.3 LEAKAGE TEST ... 36
4.1 DESIGN AN PRODUCTION OF CASINGS ... 37
4.2 HEAT SETTING OF CASINGS ... 39
4.3 EVALUATION OF BIOMECHANICAL PROPERTIES ... 39
4.4 SUSTAINABLE AND ETHICAL ASPECTS ... 42
4.4.1 SUSTAINABLE ASPECTS ... 42
4.4.2 ETHICAL ASPECTS... 43
6. FUTURE RESEARCH ... 45
7. LIST OF REFERENCE ... 46
8. APPENDIX ... 54
A. TECHNICAL SPECIFICATION OF CASING DESIGN ... 54
B. KNITTING PATTERN ... 57
C. GRAPHS OF THE AVERAGE TRANSVERSE AND AVERAGE AXIAL COMPRESSIVE FORCE WITH DECREASING DISPLACEMENT AMPLITUDE 59 D. RESULT OF THE STATISTCAL ANALYSIS FROM IBM SPSS STATISTICS 27 ... 63
1
1. INTRODUCTION
For a long period of time, the meniscus was considered to be an expendable part of the knee (Peterson & Renström, 2017), therefore total resection of the meniscus was the common procedure to treat meniscal tears. (Fairbank, 1948; Peterson et al., 2017) Nevertheless research has found that this procedure often results in retrogression of the articular cartilage and development of osteoarthritis. (Fairbank, 1948; Ghodbane, Brzezinski, et al., 2019) Since the 1980s the functionality of the meniscus and its important role in the biomechanics of the knee joint have been investigated more in depth. (Peterson et., 2017)
2 Figure 1. The vascularity of the lateral meniscus divided into its three zones: external peripheral vascular zone, the partly vascularized zone and the avascular zone.
Today there are still few meniscus replacements on the market. Two examples are the PU-based Actifit and Menaflex CMI that is based on a mixture of synthetic collagen and bovine collagen I. A disadvantage of these replacements is their unsuitable degradation rate if used as a total meniscus replacement. Menaflex CMI is considered to have too high degradation rate while the degradation rate of Actifit is too low which may cause long term residence of the implant material in the new tissue. Therefore these are mainly recommended for partial meniscectomy. (Halili et al., 2014; Vadodaria et al., 2019) Another concern is that research has shown uncertain result regarding their ability to restore the natural biomechanics of the knee joint or prevent development of osteoarthritis. (Ghodbane et al., 2019; Sun et al., 2017) Today there is one device available for total meniscus replacement called
Nusurface, which is polyethylene based with reinforcement of
polycarbonate-urethane. (Vadodaria et al., 2019) However neither of these replacements are produced using tissue engineering, therefore not capable of promoting regeneration of tissue. (Pillai, Gopinathan, Selvakumar, & Bhattacharyya, 2018)
To better mimic the internal structure of the meniscus and produce more patient specific design, bioprinting with hydrogel has been investigated as a production method for meniscus replacements. (Bandyopadhyay & Mandal, 2019) Hydrogels have advantages in tissue engineering due to their high quantity of water, biocompatibility and low mechanical and frictional irritation to the surrounding environment. (Baroli, 2007) However in general hydrogels have poor mechanical strength. (Hoffman, 2002; Baroli, 2007) For example studies have shown limited ability of collagen hydrogel to withstand load in vivo and alginate is too stiff and brittle. (Rhee, Puetzer, Mason, Reinhart-King, & Bonassar, 2016)
Due to the complexity of meniscal treatment and the limitation of available menisci replacements, research on development of efficient treatments and devices is today of increasing interest. For this reason, this thesis will focus on the possible development of a textile augmenting casing to contain, fixate and mechanically stabilize bioprinted cell seeded gels.
Peripheral vascular zone
Partly vascularized zone
3 1.1 PROBLEM DESCRIPTION
As mentioned in Section 1 there are few replacements available for meniscal reconstruction on the market. Sahlgrenska Academy, Gothenburg University has bioprinted a meniscus replacement resembling the native meniscus. This is made of a hydrogel containing alginate and nanocellulose and cell seeded with induced pluripotent stem cells (IPSCs)1 and chondrocytes2. In studies in vitro, early development of fibrous-cartilage lineage and vascular-like tubes were noted after three weeks. (Lindahl et al. 2019) Due to the limited sufficiency of biomechanical performance of a hydrogel, the bioprinted gel has limited ability to withstand tensile stresses caused by the contact compression in between the femoral and tibial articular surfaces.
Textile fibres or structures have commonly been applied in medical textiles, for example in the production of devices like vascular grafts and hernia repair mesh. (Almeida et al., 2013) The functionality of textile fibres or structures can be chemically or physically modified to meet specific requirements to enable production of medical textile with unique properties, such as absorbency, fibre length and fineness, cross-section, elasticity and stiffness. By using synthetic fibres, alterations can be made as early as in the polymerization of the fibres. An example is the specific polymer degradation rate, which can be modified by adjusting the monomer ratio. (Qin, 2015) In meniscal reconstruction, integration of textiles can work as a reinforcement to improve its biomechanical properties, resembling the structure and function of the native extracellular matrix. (Patel, Ghodbane, Brzezinski, Gatt, & Dunn, 2018) If used in tissue engineering, the textile provides a large surface for cell attachment and diffusion of nutrients. (Chen, Ushida, & Tateishi, 2002) In earlier studies, textile fibres and structures has been adopted as reinforcement in meniscal and articular cartilage replacement, for example by using techniques as weaving (Moutos, Freed, & Guilak, 2007) silk fibre composite (Stein et al., 2019) bioprinting (Bandyopadhyay & Mandal, 2019) and melt spinning (Balint, Gatt, & Dunn, 2012). Furthermore in studies by Chen et al (2003) and by Neves et al. (2006) knitted textiles were used as reinforcement in replacement of articular cartilage and menisci to enhance their biomechanical properties. In their bachelor thesis, Ivarsson & Johansson (2011) succeeded in knitting a meniscus replacement produced in Artelon filament©, but no further research was made in the project. Today, no research have been found where knitting technology has been employed in the production of a casing that is providing a bioprinted gel with biomechanical support from the outside.
1 Adult stem cells which can be reprogrammed to develop into any type of cell or tissue except from cells forming an embryo.
4 1.2 AIM
The objective for this thesis is to design and fabricate a textile casing to investigate the possibility to incorporate it to a bioprinted cell-seeded gel to enhance the necessary compressive properties for the device to properly function. The method for incorporation of the knitted casing with bioprinted gel is described in Figure 2. In simplicity, the knitted casing (A) will be produced separately on a flatbed machine with dimension to match the bioprinted gel (B) and thereafter the bioprinted gel will be put into the knitted casing (C), which will be closed by stitching before implantation. In this way, the knitted casing will provide structural support to the bioprinted gel which is necessary during reformation of the menisci. In previous studies there have been several different methods and methodologies to evaluate and test meniscus replacements. However these studies have been performed with individual methods and are difficult to compare to each other and reproduce for future work. For example indentation test was employed in a study by Abdelgaied et al., (2015) in their evaluation while Ghodebane et al (2019) used plugs of meniscus tissue in a confined compression test and tensile test in their evaluation of the biomechanical performance of the meniscus. Those studies are therefore not comparable and do not reflect a real use case of a meniscus device. Because of this there is a need for a more uniform and comparable test method that better correlates to a real usage of the meniscus.
Figure 2. An overview of the product development of a bioprinted implant with biomechanical support from a knitted casing.
A
B
5 1.3 REQUIREMENTS
This thesis is made in collaboration with Sahlgrenska Academy, University of Gothenburg. Together with them the following requirements of the casing have been determined, see Table 1.
Table 1. List of requirements for the knitted augmenting casing Requirements
1 The shape of the casing should be designed to mimic the medial meniscus
2 The casing should have reinforcements at the peripheral rim to support tensile stress and also enable device-to-bone fixation at the meniscus apex points
3 The casing design should include additional radial threads 4 The casing should be designed to withstand leakage of
hydrogel when physiologically relevant axial compression levels are applied
1.4 RESEARCH QUESTIONS
Based on the requirements from Sahlgrenska Academy the following research questions are stated to guide this thesis.
1. How can knitted casing designs add essential structural support to bioprinted cellular gels to support reformation of menisci?
The research question is divided into the following sub questions:
a) How does the amount of radial threads influence the stiffness of the knitted casing?
b) How can reinforcements at the peripheral rim of a textile casing for a meniscus replacement be achieved by knitting technology to enhance its ability to withstand tensile stresses?
c) How can a method to enable fixation of the implant be incorporated in a knitted casing design?
Since sub questions b and c are in the area of product development and choice of design they will only be answered subjectively.
6 1.5 LIMITATIONS
Development of a medical device is a complex process, often requiring several years of research and evaluation before becoming available on the market. According to ISO 14630:2012 there are general requirements to follow for non-active surgical devices for example regarding their intended performance, the design, material and sterilization etc. As an example, the biocompatibility of the construction material and possible effect of the shape and dimension of the replacement on tissue and body fluids. Since this thesis was restricted to a limited period of time, the focus was on the biomechanical performance of a first version of a meniscal replacement. Therefore the consideration of ISO 14630:2012 was limited to using biodegradable fibres utilized in clinical practice and a design suitable for a meniscus replacement.
The production and the evaluation of prototype was limited to be performed with the equipment available at the University of Borås and the Sahlgrenska Academy. Hence factors such as suitable pore size and the tribology between the bioprinted gel and surface of the knitted casing were not evaluated. Due to limited access of Artelon® filaments, the knitted casing was produced in a non-biodegradable material as a first version of the prototype to evaluate the design and determine suitable knitting parameters. This was restricted to the material already available at the Swedish School of Textiles.
7 1.6 LITERATURE REVIEW
In the literature review, the functionality of the knee joint and its menisci is introduced. Furthermore knowledge of construction of meniscal replacement and its effect on healing of meniscal tears are presented. Also bioprinting and hydrogels, knitting technology, and the Artelon® material are covered.
1.6.1 THE KNEE JOINT
8 Figure 3. Illustration of the knee joint inspired by Davis (2017) and Meyler (2018). To the left front view without patella tendon and to the right side view showing patella tendon and the synovial membrane.
The bones of the knee joint are connected by strong ligaments, which play an important role in the stability of the knee joint. The major ligaments connecting the tibia and femur are the anterior and the posterior cruciate ligaments that control the movements forward and backward, and the medial and lateral collateral ligaments that prevent the movement side to side. (Meyler 2018) Injuries of the anterior cruciate ligament can lead to excessive anterior tibia movement and are often found in combination with meniscal tears. (Zhang et al., 2016) According to Andernord et al (2014) approximately 40 % of the registered anterior cruciate ligament injuries were in association with meniscal tears. The medial meniscus is attached to the medial collateral ligament, which may result in simultaneous injuries. The knee joint is surrounded by a fibrous joint capsule. (Majumdar, 2010) The internal surface of this capsule is a thin synovial membrane, which produces the synovial fluid. This fluid lubricates the joint and transports nutrients. Furthermore it influences the compression of the joint. At bending of the joint or when load is applied, the synovial fluid is squeezed out of the cartilage, similar to a water-filled sponge. Synovial membrane also exists between the bones and the soft tissue in the form of small fluid-filled sacs called bursae. Similar to the articular cartilage, functions of the bursaes are to minimize friction. (Meyler 2018)
9
1.6.2 THE MENISCUS
In this section the functionality of the meniscus and its biochemical and biomechanical properties are described.
1.6.2.1 ANATOMY OF THE MENISCUS
In the knee joint, the menisci are positioned between the femur and tibia condyles. They have a crescent shaped geometry with a thick external circuit and a thin internal edge, see Figure 4. The lateral meniscus is round and evenly shaped and covers a larger area of the synovial surface of the tibia than the medial version. Its anterior horn fixates at the anterior cruciate ligament, while its posterior horn connects to the condyles and the back of the anterior cruciate ligament. Due to the remote fixation points, the lateral meniscus is able to move more freely back and forward than the medial meniscus. In comparison to the lateral meniscus, the medial one is more long and narrow shaped, where the width between the anterior and posterior ends differs. Its whole circuit is connected to the joint capsule and is fixated with its anterior and posterior horns to the tibia by ligamentous. (Peterson & Renström 2017) The size of the menisci are individual and influencing parameters are gender, height and weight. For example women generally have smaller menisci than men. (Stone et al., 2007)
Figure 4. A simplified illustration of the knee joint from above, inspired by Fox, Bedi, & Rodeo (2012).
Anterior horns
Posterior horns Posterior cruciate ligament
10 1.6.2.2 BIOCHEMICAL COMPOSITION OF THE MENISCUS
There are two types of cells in the meniscus: fibroblastic cells on the surface and fibrochondrocytes, which form a porous and permeable extracellular matrix (ECM). (Kawamura et al., 2003) The ECM is mainly composed of water and collagen but also smaller amounts of proteoglycans, adhesion glycoproteins and elastin. (McDevitt & Webber, 1990; Makris, Hadidi, & Athanasiou, 2011) There are various types of collagens in the different regions of the tissue, which contribute to the mechanical strength of the meniscus. In the peripheral vascular zone, collagen I is the predominating type while there is a combination of collagen I and II in the avascular zone. Proteoglycans, where aggrecan is the major group, are composed of a core protein covalent bonded to at least one glycosaminoglycan (GAG) chain. Aggrecans contribute to the viscoelastic compression properties of the meniscus tissue. Due to the negative charge and the hydrophilic properties of the molecule, it can absorb and retain water and in this way regulate the hydration of the meniscus. The purpose of the adhesion glycoproteins is to organize the ECM by intermolecularly bonding the components to the cells. The function of the fibrillary elastin is not fully verified, although it is likely to give resilience to the tissue. (Makris et al., 2011; Fox, Bedi, & Rodeo, 2012)
1.6.2.3 BIOMECHANICAL PROPERTIES OF THE MENISCUS
The menisci play an important role in the biomechanical function of the knee joint, where their primary task is to absorb mechanical shock, transmit load and distribute compressional stress between the femoral and tibia bones. Furthermore the menisci provide stability and joint lubrication. (Sweigart & Athanasiou, 2001; Fox et al., 2012)
11
The meniscus has anisotropic mechanical properties, which is a result of its complex ECM structure. (Ghodbane, Patel, et al., 2019) The orientation of the collagen in the meniscus can be divided into three different layers, see Figure 5. In the top layer, called the superficial, the collagen fibres are oriented without any in-plane order. This random order also appears in the lamellar or middle layer, except from the radially oriented fibres in the peripheral of the anterior and posterior horns. The fibres in the deep layer are circumferentially oriented with a small amount of radial tie fibres. (Sweigart & Athanasiou, 2001) This distinct fibre structure contributes to the anisotropic properties in tension, compression and shear resistance of the meniscus. (Fithian, Kelly, & Mow, 1990) In a study of the tensile stress properties of the meniscus by Tissakht & Ahmed (1995), the result showed that the collagen fibres in circumferential direction are stiffer and are able to withstand higher stress in comparison to the radial fibres. In addition, the tissue appeared to be less stiff in the middle layer compared to the superficial and the deep layer. By indentation testing, Danso et al. (2015) found the medial meniscus more site dependent on its mechanical properties in comparison to the lateral one.
Figure 5. The layers of the collagen and its orientation in the meniscus. Sketch inspired by Sweigart & Athanasiou, 2001.
The shear strength of the meniscus tissue has an intrinsic viscoelastic performance, which depends on frequency, shear and compressive strain. (Zhu, Chern, & Mow, 1994) The viscoelastic compressive properties depend upon permeability of the meniscus. A higher amount of proteoglycans will increase the absorption of water, thereby improving the compressive properties by lowering the shock. (Kawamura et al., 2003; Ghodbane, Brzezinski, et al., 2019) When an axial compressive load is applied to the knee joint, the menisci articulate on the surface of the tibia and femora. Due to their wedge-shaped geometry, the menisci increase the contact area of the femoral condyle and tibial plateau, reducing the contact stress which can appear upon the articular cartilage and enabling transmission of compressional load. (Ghodbane, Patel, et al., 2019) Previous studies have found the average movement of the medial meniscus under load to be 3.6 mm (Vedi et al., 1999) or 2.6 mm +- 1.2 mm (Freutel et al., 2014). In the posterior and the anterior horns the translations are 6-7 mm and 3-4 mm respectively. (Yao, Lancianese, Hovinga, Lee, & Lerner,
12
2008; van Eisenhart-Rothe, et al., 2004) However the displacement of the menisci are also influenced by the degree of flexion of the knee joint. A larger flexion of the joint reduces the tibiofemoral contact area, causing larger displacements to the menisci. (Sweigart & Athanasiou, 2001) Compression of the menisci also influences the lubrication of the knee joint. When compressed, synovial fluids are transferred into the articular cartilage to reduce the friction and to transport nutrients into the avascular zone. (Fox et al., 2012).
1.6.3 MENISCAL INJURIES
Meniscal tears are the most common intra-articular injury to the knee. (Abrams et al., 2013; Halili, Hasirci, & Hasirci, 2014) The injuries can be divided into the two categories: traumatic and degenerative. (Bozkurt, Unlu, Cay, Apaydin, & Dogan, 2014; Peterson et al., 2017) To be able to develop a suitable design for the knitted casing, the common tearing pattern had to be considered to understand where additional textile reinforcement were required.
13 Figure 6. Overview of common meniscal tears. A: Healthy meniscus, B: Longitudinal, C: Horizontal, D: Radial, E: Parrot/Oblique, F: Root, G: Flap, H: Bucket handle and I: Complex or Degenerative. Sketch inspired by Lecouvet et al. (2018), Karia, Ghaly, Al-Hadithy. N., Mordecai, S., & Gupte, C. (2019).
1.6.4 CONSTRUCTION OF A MENISCUS REPLACEMENT
Construction of a meniscus replacement is a complex process with design on both macro- and micro-molecular level. In this section the subject of tissue engineering and relevant factors to consider in development of a meniscus replacement are introduced.
Tissue engineering is a regenerative treatment of lesions including living cells. In this method the cells are often combined with a scaffold of biomaterial, which aim to provide structural support in reconstruction of new tissue. (Chen et al., 2002; Szojka et al., 2017) It is a repeatable process which allows reproduction of tissue with control of design and enable flexible material combinations to achieve suitable degradation rate and mechanical properties of the device. (Vadodaria et al., 2019) When producing a meniscus replacement for meniscus, important parameters are choice of construction material, shape stability in vivo, porosity and fixation of device. The chosen material should be nontoxic and non-carcinogenic. Otherwise a negative immune response in vivo with complications like inflammation, destruction of host tissue and rejection of the implant may appear. (Vadodaria et al. 2019) In addition the material should be able to adhere to the native tissue to avoid gap formation under load which may interrupt the healing. (Vishwakarma et al., 2016; van Tienen et al., 2002; Halili et al., 2014) Moreover, regarding tissue engineering, the material is required to allow cell adhesion onto the surface and enable migration and proliferation of the cells. (Pillai et al., 2018) The choice of material also influences the biomechanical performance of the meniscus replacement and its shape stability. Preferably the compressive and tensile properties of the meniscus replacement should resemble those of the native tissue at the time of the implantation. (de Groot et al., 1997; Hutmacher, 2001; Bat et al.,
14
2014; Zaffagnini et al., 2015) If the device is too stiff, it may cause stress on the surrounding cartilage, while too stretchable is incapable of giving support to the knee joint and the implant may collapse or shrink. (Zaffagnini et al., 2015; de Groot et al., 1997) As an example, a study by Sandmann et al. (2013) showed that CMI and Actifit possessed 26-27% of the compressive properties of the native meniscus. A close match of biomechanical properties also induces the formation of fibrocartilage. (de Groot et al., 1997) When constructing a biodegradable meniscus replacement, the degradation time of the material needs to be taken into consideration. The degradation rate of the material should match that of the tissue regeneration. (Liu, Thomopoulos, & Xia, 2012; Hutmacher, 2001) This is because the degradation of the material affects the compression and density of the device, thereby reducing its mechanical properties. Therefore the degradation process should start after the induction time of the fibrocartilage. (van Tienen et al., 2002)
The hydration rate of the biomaterial is affected by type of bond, presence of water or enzymes, glass transition temperature, hydrophilicity, morphology, crosslinking and pH. (Bat et al., 2014)
The effect on the ingrowth of tissue and healing capacity of the meniscus regeneration depends on the porosity, pore size and the compression modulus of the replacement. (van Tienen et al., 2002) In their study, it was found that a meniscus replacement with larger porosity and pore size had a complete ingrowth of tissue before the degradation of the polymer in comparison to smaller ones. According to Yan et al., (2012) higher porosity resulted in a higher water uptake ratio, which affects the swelling of the scaffold. In addition pores in macro size transport the nutrients and metabolic waste. Meanwhile the micro pores increase the efficiency of the cell seeding, enhance the cell adhesion and degradation of the implant. When fixing the meniscus replacement the surgical technique should be as gentle as possible to prevent damage of the articular cartilage. Furthermore the fixation should be rigid enough to withstand the forces and generation of tensile stresses applied to the implant, while allowing the pattern of movement in flexion and extension of the knee joint. (Rongen, van Tienen, van Bochove, Grijpma, & Buma, 2014)
1.6.5 EVALUATION METHODS FOR BIOMECHANICAL PERFORMANCE OF MENISCUS REPLACEMENTS
15
conducted tests have described the material and anisotropic nature of the meniscus tissues rather than the biomechanical characteristics of the menisci or its substitute as a whole. By adoption of indentation test it is possible to measure the elastic modulus in a non-destructive and site-specific way. (Danso et al., 2015) In an evaluation of testing of compression tests on articular cartilage by Korhonen et al. (2002) the elastic modulus was found to be higher in indentation test in comparison to confined and unconfined compression tests, which may be related to the damage of the fibre structure in the tissue due to the punching of samples.
1.6.6 BIOPRINTING AND HYDROGEL
Bioprinting, enables deliverance of cells within the bio-ink. This can be used to create a three-dimensional defined structure of a CAD model. The medical image which the CAD model is based on can be generated from techniques such as CT scanning and MRI. (Hospodiuk, Dey, Sosnoski, & Ozbolat, 2017) Moreover is allows production of gels with patient specific design and the possibility to build internal architecture on micro-level, which is a limitation of the traditional casing technologies. Other advantages are that it is a fast and repeatable technique. (Bandyopadhyay & Mandal, 2019) By using bioprinting in the development of a meniscus replacement, it is possible to control the fibre diameter, fibre orientation and the pore structure in progress of creating a replacement mimicking the ECM structure of the native meniscus. (Sun et al., 2017; Murphy & Atala, 2014)
Initially, in 3D printing polymer, metals and ceramics was used as printing materials which often requires high temperatures and additives like crosslinking agents and organic solvents. Nevertheless, bioprinting is a more sensitive process where temperature and additives may harm the cells in the biomaterial reducing their bioactivity. (Murphy & Atala, 2014; Hung, Tseng, Dai, & Hsu, 2016) In the choice of printing material, important parameters to consider are printability, biocompatibility, and structural and mechanical properties. Furthermore the material needs to have proper rheology and thermal properties to depose the correct amount of material to achieve a satisfying design result. (Narayanan et al., 2016) Regarding the cellcompatibility and regeneration of tissue, the material is required to have proper degradation kinetics, to be nontoxic and non-immunogenic and promote cell adhesion.
16
are crosslinked with molecular entanglements or outside forces are referred to as reversible hydrogels. Ionic crosslinked hydrogels are one such reversible hydrogel. Reversible or more stable characteristics can in some cases be a driving factor for choice of crosslinking method. In medical purposes reversible if often preferred since it results in a decomposable hydrogel. (Hospodiuk et al., 2017)
1.6.7 ARTELON®
The purpose of biodegradable materials in medical devices is to first induce ingrowth of new tissue, thereafter undergo controlled decomposition into non-cytotoxic substances in vivo. (Guelcher, 2008) Artelon® is made of poly-(urethane urea)s (PUU), which is a multiblock copolymer with an elastic modulus above the rubber level. The phase-separated heterogeneous structure of the PUU is divided into soft and hard segments. The amount of hydrogen bonded hard blocks in the soft matrix influences the mechanical strength, modulus and elasticity of the material and enables a variety of material properties depending on the application. (Gisselfält, Edberg, & Flodin, 2002) After surgery, the degradation rate of the Artelon® is four to six years, while during this time new tissue is regenerated to replace the implant material. (Artelon 2020)
Artelon® is synthesised by using block polymerization in two steps. In the first step polyester diol PCL530 is slowly added to 4.4´-diphenylmethane diisocyanate (MDI), at 50 ˚C in a dry N2environment to form a prepolymer. In the second step
a diamine chain extender and monoamine chain-stopper, diluted in N, N-dimetylformamid (DMF) solution, is quickly added during intense stirring to a solution of prepolymer and DMF at 20 ̊C. PUU filaments are produced using wet spinning. The prepared polymer solution is extruded through a spinneret into a coagulation bath with water that extracts the solvent. Next the fibre bundles are drawn into the preferred draw ratio. In the last step the produced filaments are reeled up on spools and rinsed in water before drying. (Gisselfält et al., 2002)
1.6.8 KNITTING TECHNOLOGY
17
extensible and lighter fabric, but lowers the bursting strength. (Spencer, 2001) Nevertheless knitting technology has drawbacks including poor dimension stability and high shrinkage after the production due to high tension in the formation of loops. (Au, 2011; Ray, 2012)
2. EXPERIMENTAL SECTION
In the experimental section the material and the methods applied in the development of the meniscus casings are presented. Furthermore characterization of the casings and the statistical analysis are described.
2.1 MATERIAL
In this section the materials used in the fabrication of the casings and in the production of prototypes of the bioprinted gel are presented.
2.1.1 CASING MATERIAL
18
2.1.2 BIOPRINTED GEL AND PROTOTYPES
A bioprinted gel without cell-seeding was produced as a reference for the gel prototypes by Sahlgrenska Academy, see Figure 7. It was made of nanocellulose and alginate hydrogel with a 60/40 dry weight % ratio and cross-linked in 90 mM CaCl2.
Figure 7. Bioprinted gel with the shape of a meniscus from a specific patient on a 9 cm diameter petri dish.
19 2.2 FABRICATION METHOD
Firstly in this section, the fabrication of the bioprinted gel prototypes used in the production and the evaluation is presented. Secondly, the development process of the augmenting casing are described. In short the process included following steps: design, knitting and heat setting.
2.2.1 BIOPRINTED GEL PROTOTYPES
To produce testing samples for the evaluation of the test method and for the heat setting, a mould of the bioprinted gel was made of plaster gauze and water. The mould was dried at room temperature for an hour before removing the bioprinted gel from the casket, then the mould was dried for additional three hours to become completely dry.
With the plaster mould silicone prototypes of the bioprinted gel were produced, see Figure 8: A. These were later used as a template in the measurement of correct dimension of the casing during the knitting process and for the heat setting of the knitted casings. 30 g of silicone paste was prepared according to the producer’s instruction, and put in the mould. Thereafter the prototype was air dried in room temperature for five minutes.
Figure 8. Example of the produced implant prototypes A: silicone and B: gelatine.
As mentioned in section 1.6.2.2, the ECM of the native meniscus is mainly composed of water and collagen. (McDevitt & Webber, 1990; Makris, Hadidi, & Athanasiou, 2011) According to Fox et al., (2012) the amount of collagen in the ECM is about 22%. Therefore after testing, a concentration 22% gelatine was chosen for the moulding of the gelatine gels. In the first step, 22 g of gelatine powder was stirred into 100 ml of water. Thereafter the solution was allowed to sit for 5 minutes before heating up to 45℃. Afterwards 20 ml gelatine solution was added into each plaster mould and put into the refrigerator at 11 ℃ for the prototype to gel for at least 4h.
20
2.2.2 DESIGN AND KNITTING OF CASINGS
The choice of design of the augmenting casing was based on the set requirements, see section 1.2. Since the size of the meniscus is individual depending on parameters such as gender, height and weight (Stone et al., 2007), the dimension of the knitted casing was made to fit one specific bioprinted gel. To evaluate the influence in stiffness by the amount of radial threads of in the augmenting casing, two different casing designs were produced with varying amount of additional radial threads. The design of the casing was an iterative process, where the knitting pattern was developed step-wise by evaluating knitted samples on the silicone prototype.
Knitting was chosen as a production method due to its possibility of direct manufacturing of products, flexibility in choice of design, minimal waste of material and the likelihood of introducing weft inlays in the knitting pattern. (Ray, 2012) Furthermore the multidirectional stretchability (Au, 2011; Ray, 2012) made it possible to design a casing with a close fit without additional folding of the fabric. Firstly, knitting patterns for the casings were designed in Stoll M1plus (Reutlingen, Germany). Thereafter the two casing designs were produced on an ADF 530K flatbed knitting machine from Stoll (Reutlingen, Germany) which had the machine gauge 14 and needle size 12. Due to limitation of material supply and to minimize the required amount of Artelon® filament to reduce the project cost, first prototypes were made in polyester yarn to determine machine parameters and to create a suitable knitting pattern for the casings. The polyester casings were knitted with one yarn in the yarn carrier. Furthermore full rib ribbons were knitted to mimic the Artelon® Flexband™ in the evaluation of the characterization method of the compressional axial behaviour of the casing. To avoid breakage of filaments, the Artelon® was reeled onto new cones before the knitting process. In the knitting of the Artelon® prototypes two filaments were used in the yarn carrier both for the knitting of the casing shape and for the weft inlays. This to match the yarn density of the polyester yarn which the base pattern was constructed with. To minimize friction between the filaments, fournisseurs were used with two filaments in each. For the weft inlay, EFS 920 from Memminger-IRO (Dornstetten, Germany) was used to achieve a positive feeding of the weft yarn.
2.2.3 HEAT SETTING OF CASINGS
21 2.2.3.1 HEAT SETTING OF POLYESTER CASINGS
The heat setting of the polyester casings was performed with an oven of model G209 from SDL Atlas (Rock Hill, SC, United States). Before the treatment, the
knitted casings were relaxed for at least 12 h to remove the strains caused by the knitting process. The weft threads were stretched and evenly distributed through the knitted structure. Thereafter the casing was hand stitched onto a metal casing of steel wire and heat treated in 160˚C for 7 minutes. To avoid additional shrinkage of the structure after the heat setting, the knitted sample was first removed from the metal prototype when completely cooled down.
2.2.3.2 HEAT SETTING OF ARTELON CASINGS
From the result of the evaluation of the heat setting of the Artelon®, the heat gun was found the most suitable. Before the treatment, the knitted casings were relaxed for at least 12 h to remove the strains caused by the knitting process. The heat setting was performed using a Cotech heat gun from Clas Ohlsson (Insjön, Sweden) and the silicone prototype. Before the heat setting, the casing was soaked in cold water to make it more flexible and easier to put onto the silicone prototype. The casing was hand stitched onto the silicone prototype and treated at 100˚C for approximately 10 minutes to achieve a close fit. Just as with the polyester prototypes, the Artelon® casings were cooled down on the silicone prototype.
2.3 CHARACTERIZATION METHODS
In this section the chosen methods to evaluate the axial compressional force and the tensile stiffness of the knitted casings and testing of possible leakage of gel material are presented.
2.3.1 VALIDATION AND FRAME OF REFERENCE TESTING
22
2.3.2 EVALUATION OF AXIAL COMPRESSIONAL BEHAVIOUR
As mentioned earlier in Section 1.6.5, there is still no standardized test method to evaluate the compression or tensile properties of meniscus replacements. Standard ASTM F3223-17 only recommends properties to measure, but not how the test should be performed. Earlier work has reported testing with punched samples. Nevertheless, to be able to find an answer to the set of research questions in this project, the evaluation of the axial compressional behaviour were performed on complete knitted casings with the moulded gelatine gels inside.
Figure 9. Illustration of the testing advice used to evaluate the compressional behaviour induced by cyclic tensile loading.
The experiment was performed with a hysteresis test using a tensile tester Tensolab 2512A/2512C from Mesdan (Brescia, Italy) to replicate the loading of force to the knee joint. For this a special device with a four cm diameter metal ball welded onto a flat metal plate was manufactured and fitted to the tensile tester to simulate the compressive pressure applied on the meniscus from the femorotibial joint, see Figure 9. In this way stretching of the gripped device wedged it into the gap between the ball and the plate in an inverse manner to the anatomical mode of action. Thereby mimicking the anatomical case where the weight transferred from the femur condyle will wedge the medial meniscus medially (and the lateral meniscus laterally). By measuring the force in the fixation ribbon ends and analysing how the geometry generates contact and frictional forces the leverage perpendicularly can
F Sample
Additional device
15mm gap between upper clamp and sample tip
23
be managed and the impact of femorotibial loading can be assessed. The test was performed in a conditioned lab with room temperature 20 ± 2˚C and humidity 65 ± 2%. The set parameters are presented in Table 2. The number of specimens was five per polyester casing design and four per Artelon® casing design.
To reproduce the in vitro environment the casing will be exposed to when implanted, the samples were soaked in deionized water for 20 minutes. To reduce the friction between the textile and the metal surface, the underside of the specimen and the metal plate were greased with Chesbrough Kløver Vaseline® Unilever (Rotterdam, Netherlands). A sample was mounted around the metal ball and fixated at the ribbons using pneumatic clamps as illustrated in Figure 8. The cyclic tests were performed once per specimen. Afterwards the mean values and standard deviation of the forces retrieved was determined using Microsoft Excel 2013 to be used to calculate the axial compressional force, see Section 2.3.2.
Table 2.The set parameters for the hysteresis test. Parameter
Load cell [kN] 0.1 Gauge length [mm] 131* Pretension [cN] 55±2 Loading rate[mm/min] 800 Maximum displacement
amplitude[mm]
Cycle 1 Cycle 2 Cycle 3
3 5 7
*Note that this value is measured between the two clamps, the distance
24
2.3.3 CALCULATION OF AXIAL COMPRESSIONAL FORCE
The transverse force (F) retrieved from the hysteresis test, was used to calculate the axial force applied to the tested sample. This was performed using the forces described in Figure 10.
Figure 10. Illustration of the distribution of force that act on the sample during the hysteresis test. F is the transverse force measured by the load cell of the tensile test with the opposing frictional force, Fax, μ according to Newton’s third law. Fax is the compressional force between the metal plate and the metal ball, N is the normal force and Nμ the resultant frictional force applied to the sample.
Force balance along the X-axis renders equation (1) and force balance in the Y-axis direction renders equation (2).
𝐹 − 𝐹𝑎𝑥𝜇 − 𝑁𝜇𝑐𝑜𝑠(𝛼) + 𝑁𝑠𝑖𝑛(𝛼) = 0, (1)
𝑁𝑐𝑜𝑠(𝛼) + 𝑁𝜇𝑠𝑖𝑛(𝛼) − 𝐹𝑎𝑥 = 0, (2)
25
Firstly the average angle α was calculated, where the angle was measured where the largest force applied to the sample was estimated see Figure 10. In simplicity α was approximated by the ball tangent at the contact area with the sample. Sample height and radial distance formed the catheters. Then the angle was given by:
𝛼 = 90˚ − tan−1 𝑥
𝑦 (3)
Equations (1) and (2) were altered and set to equal to determine an expression for the axial compressional force Fax , equation (4).
𝐹𝑎𝑥= 𝐹 (
1 𝜇−𝑠𝑖𝑛 (𝛼) −𝜇𝑐𝑜𝑠 (𝛼)
𝑐𝑜𝑠 (𝛼) +𝜇𝑠𝑖𝑛 (𝛼)
) (4)
2.3.4 DETERMINATION OF COEFFICIENT OF FRICTION
The aim of the experiment was to determine the coefficient of friction (COF) generated by the frictional force between the lubricated textile surface of the knitted casing and the lubricated metal plate in the evaluation of stiffness, see equation (1). This was necessary since the COF is a variable in the calculation of the axial force. To consider possible statistical differences in COF between the two materials and their influence in the result in the calculation of the axial force, the COF was determined both for the polyester and the Artelon® casings.For the calculation of the COF Coulomb’s law of friction was applied:
𝐹𝑓𝑟= 𝑁𝜇 (6)
Since the sample is at rest at the start of the hysteresis test, both the static coefficient μs and the kinetic coefficient μk was determined, where the μk was calculated when
force was considered constant. The normal force in (6) is given by:
𝑁 = 𝑚𝑔 (7)
26
The frictional force was determined by a friction test using tester Tensolab 2512A/2512C from Mesdan (Brescia, Italy) together with an additional pulley system, see Figure 11 below. A similar method has earlier been applied by Hermann, Ramkumar, Seshaiyer, & Parameswaran (2004) in their study of friction and woven fabrics.
Figure 11. Sketch of tensile tester with additional pulley system for friction test.
The metal plate was attached on a wooden plate and greased with Vaseline ®. To mimic the in vitro environment the casing is supposed to be working in when implanted, the samples were tested in wet condition. Therefore the evaluated samples were soaked in deionized water for 20 minutes before testing. A polyester thread was sewn into the end of the full rib ribbon/Flexband™ of the knitted sample and its underside was greased with Vaseline®. Thereafter the sample was put onto the metal plate and the additional polyester threads were fixated with pneumatic grips in the tensile tester. A tensile testwas run, which set the sample into moving an on beforehand determined distance. The numbers of samples was one per casing design and number of testing cycles ten per sample. The test was performed in a condition lab with room temperature 20 ± 2˚C and humidity 65 ± 2 %. In Table 3, the set parameters for the friction test are presented. The average values, calculation of the COF and the standard deviations were calculated in Microsoft Excel 2013.
Additional load Sample
27 Table 3. The set parameters for the test of frictional force.
Parameter
Load cell [kN] 0.1 Gauge length [mm] 100 Loading rate [mm/min] 50 Distance [mm] 30 Additional load [kg] 0.598
2.3.5 LEAKAGE TEST
To fulfil the requirement of the casing to withstand leakage of hydrogel when exposed to axial compressional force, a leakage test was performed on the gelatine prototypes before and after the hysteresis test. This by visually inspecting and document possible splits or other damage to the sample. Deformation of the implant due to the compression in the test was not considered as damage to the implant in the case of leakage.
2.4 STATISTICAL METHODS
To investigate the influence of the amount of the radial threads in the two casing designs in regard of their axial compressive response, the result of the characterization of the casings was statistically analysed. Firstly to determine if the data set was normally distributed, a Shapiro-Wilk’s test was performed. The Shapiro-Wilk’s test was chosen for the normality test since it is often applied in statistical analysis with a small sample size (n<50) (Mishra et al., 2019), meanwhile it provides with a better power in comparison to for example Kolmogorov-Smirnov test (Ghasemi & Zahediasl, 2012). Due to the linear relationship between the transverse force (F) and the axial compressional force Fax, see equation (4), only the
data set for the transverse force was analysed in the normality test. The data set of the transverse force was retrieved from the hysteresis test. In the analysis the following null hypothesis was applied:
H0: The data set of the transverse force F is normally distributed. H1: The data set of the transverse force F is not normally distributed.
Since the focus of the project was to investigate the influence of radial threads in the design of the casing, there was no statistical comparison between the two materials. To statistically differentiate the two casing designs in their axial compressive response, the collected data was evaluated using a one-way ANOVA. The following null hypothesis was applied for the statistical analysis:
H0: There is no statistical significant difference between the two casing designs. H1: There is a statistical significant difference between the two casing designs.
28
et al., 2019; Stein et al., 2019), the significance levels was set to p < 0.05 for all the testing.
3. RESULT
In the following section, the results of the production of the casings and the evaluation of the characterization are presented.
3.1 FABRICATION METHODS
This section presents the results of the design, knitting and heat setting of the polyester and Artelon® casings.
3.1.1 DESIGN AND KNITTING OF CASINGS
The result of the design of the casings is presented in Figure 12 and the knitted casings with an example of the pattern in Figure 13. A technical specification of both casings is presented in Appendix A. The knitting pattern, described more in detail in Appendix B, can be divided into three sections, A-C. In section A and C interlock stitch was made with stitch length 8.5 and in section B a tubular shape was made of plain stitch with stitch length 11. The purpose of the tube was to create a channel for the insertion of a Flexband™. In this way, the requirement of reinforcement at the peripheral of the casing and fixation could be achieved. To achieve a circular and crescent-shaped design, partial knit was included in the pattern with varying number of needles in work. A purl stitch was made between section A-B and B-C, where the pattern changed from interlock to tubular plain
stitch to create a pleat effect. In this way, a folding effect was made resulting in a 3D effect at the peripheral rim of the casing design.
Figure 12. Design of knitted casings. To the left, option 1 with weft insertion in every pattern repeat and option 2 to the right.
Opening for insertion of bioprinted implant Channel for insertion of Flexband™ Weft inlay
29
The two designs have the same dimensions, although differ in the amount of radial reinforcements, which was achieved by weft inlays. The differences are primarily in the partial knitting. In the polyester casings, design 1 had 10 inlays in the partial knit in total 22, while design 2 had 6 inlays in the partial knit and a total of 18. This resulted in a 40 % difference of radial reinforcement between the two designs in the partial part of the pattern and in total 18%. Due to the difference in yarn density and amount of shrinkage between the two materials, the knitting pattern was adjusted when knitting in the Artelon®. In these patterns, the width was increased with two stitches in course direction and four repeats of partial knit were added in wale direction in the middle of the pattern to increase the length. However, the addition of sections of partial knit also influenced the number of weft inlays, where design 1 had 14 weft inlays and in total 26 and design 2 had 8 inlays and a total of 20. The difference of radial reinforcement in the two Artelon® designs were approximately 42.9 % in the partial section and in total 23.1 %.
Figure 13. The knitted Artelon® casing design 1 and section its knitting pattern to the right.
30
3.1.2 HEAT SETTING OF CASINGS
To control the dimension stability of the knitted structure and to enhance the stiffness of the fibres the knitted casings were heat set. An example of a heat set Artelon® casing is shown in Figure 14.
Figure 14. Artelon® casing heat set by heat gun.
In Table 4, the average shrinkage of the heat set casings are presented. The shrinkage of the circumferences were measured in course direction and the width of the mid, anterior and posterior areas were measured in wale direction of the knit, see Appendix A.
Table 4. Average shrinkage in percent of heat set knitted casings, the specified
measuring points are included in the technical specification.
31 3.2 CHARACTERIZATION AND STATISTICAL ANALYSIS
In this section the result of the characterization including statistical analysis of the knitted casings is presented.
3.2.1 EVALUATION OF COEFFICIENT OF FRICTION
In Figure 15, the frictional force as a function of displacement is presented for the polyester and the Artelon® samples. The average static COF was calculated from the maximal friction force, which can be noted as the peak in the graph and the kinetic COF from where the force can be assumed to be constant.
Figure 15.The average frictional forces for polyester and Artelon® casings.
In Table 5 below, the calculated COF for polyester and Artelon© casings are presented. For the determination of the kinetic COF, values from 28-33.6 mm was used for the polyester and 20-25 mm for the Artelon©.
Table 5. The average static and kinetic COF of the polyester and Artelon© casings.
Sample Average static COF Average kinetic COF
32
3.2.2 EVALUATION OF AXIAL COMPRESSIONAL BEHAVIOUR
In the following section the result of evaluation of the axial compressional behaviour is presented. The results are calculated using the kinetic COF, which was also used in the statistical calculations. The graphs are based on the force with increasing displacement amplitude of the hysteresis test, the graphs with decreasing displacement amplitude and the result of casing design 2 are presented in Appendix C.
In Figure 16, the result of the average transverse forces F in the polyester casing is presented. During the cyclic test, the highest force measured was 3.74 N in the third cycle with displacement 7 mm.
Figure 16. The result of the average transverse force of the polyester casing design 1 with incrementally increasing displacement. Cycle 1=3 mm displacement, Cycle 2=5 mm displacement and Cycle 3=7 mm displacement.
Below, the result of the average transverse forces F in the Artelon® casing design 1 are presented see Figure 17. In graph, the highest force measured 5.86 N can be noted at displacement 7 mm. 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 0 1 2 3 4 5 6 7 8 F [N] Displacement [mm]
Average transverse force polyester design 1
33 Figure 17.The result of the average transverse force of the Artelon® casing design 1 with
incrementally increasing displacement. Cycle 1=3 mm displacement, Cycle 2=5 mm displacement and Cycle 3=7 mm displacement.
The result of the Shapiro-Wilk’s test showed that the data set from the hysteresis test was normally distributed since the overall significance values of the data set was greater than 0.05. An example of the significance levels for design 1 for the polyester and Artelon® casings with maximum displacement 7 mm is presented in Table 6. For all results, see Appendix D.
Table 6. The result of the Shapiro-Wilk’s test
of the Polyester and Artelon casing design 1.
Polyester Artelon® Displacement
[mm]
Sig. level Sig. level
0 0.361 0.612 0.5 0.734 0.065 1.00 0.694 0.163 1.50 0.876 0.498 2.00 0.342 0.381 2.50 0.940 0.395 3.00 0.988 0.220 3.50 0.508 0.796 4.00 0.963 0.347 4.50 0.942 0.202 5.00 0.917 0.894 5.50 0.939 0.538 6.00 0.903 0.618 6.50 0.080 0.261 7.00 0.999 0.745 0 1 2 3 4 5 6 7 0 1 2 3 4 5 6 7 8 F [N] Displacement [mm]
Average transverse force Artelon design 1
34
Since the data of the transverse force could be regarded as normally distributed, it could be used to calculate the axial compressional force of the two casing designs and statistically analyse the result with a one way ANOVA. The average height of the gelatine gels resulted in an average angle α= 29.9˚, which was employed for the calculation of the axial compressional force of the polyester and Artelon® casings. By equation (4), the average axial compressional forces on the polyester casing was calculated, see Figure 18. In the result the highest force achieved was 29.77 N at 7 mm displacement.
Figure 18. The result of the calculated average axial compressive forces of the polyester casing design 1 with incrementally increasing displacement. Cycle 1=3 mm displacement, Cycle 2=5 mm displacement and Cycle 3=7 mm displacement.
From the calculation of the axial compressive forces of the Artelon® using equation (4), the highest value retrieved was 10.82 N in cycle 3, see Figure 19.
Figure 19. The result of the calculated average axial compressive forces of the Artelon® casing design 1 in function of incrementally increasing displacement. Cycle 1=3 mm displacement, Cycle 2=5 mm displacement and Cycle 3=7 mm displacement.
0 5 10 15 20 25 30 35 0 1 2 3 4 5 6 7 8 Fax [N] Displacement [mm]
Average axial compressional behaviour
Polyester design 1
Cycle 1 Cycle 2 Cycle 3 0 2 4 6 8 10 12 0 1 2 3 4 5 6 7 8 Fax [N] Displacement [mm]Average axial compressional behaviour
Artelon design 1
35
The result of the statistical analysis showed no statistical difference between the two designs, since the p-value for all three cycles were greater than 0.05, see Table 6. For the complete analysis, see Appendix D.
Table 7. The p-value of the polyester and the Artelon® casing designs at the three
cycles. Displacement [mm] Polyester p-value Artelon® p-value 3 0.937 0.994 5 0.538 0.986 7 0.688 0.973
In Figure 20, a comparison of cycle 3 of the two casing materials including their two designs are presented. The result of the calculation of the axial compressional forces for cycle 3 is presented in Appendix E. From the derivative of the curves, it can be noted that the polyester had a higher stiffness against axial compression compared to the Artelon®.
Figure 20. A comparison of the average axial compressional forces in function of displacement for the polyester and Artelon® casings for cycle 3 = 7 mm.
0 5 10 15 20 25 30 35 0 1 2 3 4 5 6 7 8 Fax[N] Displacement [mm]
Average axial compressional behaviour cycle 3
36
3.2.3 LEAKAGE TEST
In the visual evaluation of the gelatine gels after the hysteresis test, no damage or splits could be detected among the inspected samples.
Figure 21. Samples of gelatine gels before (A) and after (B) the hysteresis test.
37
4. DISCUSSION
In this section the production of the casings and the evaluation methods are discussed. Moreover a reflection of the sustainable and ethical aspects of the project are presented.
4.1 DESIGN AN PRODUCTION OF CASINGS
In this project knitted augmenting casings have been developed to contain and mechanically stabilize cell seeded hydrogels. In the development of meniscus replacement a parameter to consider is the construction material (Vadodaria et al. 2019) and the degradation rate of the chosen material, which should match the one of the tissue regeneration (Liu et al., 2012; Hutmacher, 2001). As casing material, Artelon® filaments were used, which has been approved for clinically application and possess a degradation rate of four to six years. Hopefully, the cells in the bioprinted hydrogel can regenerate new meniscus tissue during this time, while supported by the augmenting knitted casing.
The design of the casing was based on the requirements set in Section 1.2, with design features as reinforcement at the peripheral rim of the casing, additional radial threads and enabling device-to-bone fixation. The casing was designed to mimic the medial meniscus and formed to fit the dimension of one specific bioprinted gel, this because of the individual dimensions of the native meniscus. (Stone et al., 2007) The production of the casing was limited to the available flatbed machines at the Swedish School of Textiles and therefore to a specific machine gauge. The restriction of machine gauge, limited the flexibility of the size of the knitted casing. This resulted in samples that were slightly too large and too sharp at the tip of the horns. To achieve a better fit to the hydrogel, the knitting pattern needs to be modified to achieve slimmer and more circular horns of the casing. The knitting pattern should also be adjusted to reduce the dimension of the casing to enable a broader variation of size.
38
When axial compressional forces are applied to the knee, these are converted into tensile stresses in circumferential direction of the surface of the meniscus. (Fox et al., 2012) In meniscal injuries, the tearing often appear at the internal circumference in the avascular zone of the meniscus, see Figure 5 in Section 1.5.3. To achieve a crescent shaped casing and because of the machine gauge, there were few needles in work at the internal circumference of the casing. This could be a possible weak point of the structure when high forces are applied. With a higher machine gauge, there would be possible to introduce more needles in work, thereby adding more stitches to the structure that can provide with dimension stability and strength. In the future it may also be necessary to introduce extra reinforcement at the internal circumference to improve the strength where the risk of ruptures is most profound. This design feature was not possible with the current developed knitting pattern. It may be possible to introduce a channel in the inner circumference adjusting the interlock in the knitting pattern but this requires further development and evaluation. Furthermore, it is important to perform the adjustments of the knitting pattern without interfering with the other design requirements of the casing.
The requirement of additional radial threads was achieved by inclusion of weft inlays in the knitting pattern. The total amount of and the possible difference in weft inlays between the two casings were limited by the machine gauge. Nevertheless the result of the statistical analysis showed no significant difference between the two casing designs. In the result, thep-value was especially high when comparing the two designs produced in Artelon® in all three cycles. However, due to the small delta in the averages of the two designs, see Appendix E, the result of the one-way ANOVA was fairly expected. In the characterization of the casing designs, the polyester samples showed more variation in the result if compared to the Artelon®. It should be noted that the polyester samples were evaluated earlier on in the project, therefore the influence of practice mounting the samples could have affected the result causing a larger variation in the test data. However, type of material in the sample could also be a possible influencing parameter of the result. Nevertheless to verify or deny this requires further testing specifically designed for this problem. From the result of the statistical analysis it may be assumed that number of the additional radial threads in the knitted structure had a low impact on the stiffness and the axial compressive response of the casing. For this reason, instead of inclusion of extra radial support, the design of the casing should be designed with the focus on improvement of the tangential support since this is the direction where the largest amount of tensile stress is applied.
