These techniques are frequently used to treat osteochondral defects of high diameter. They use transplantation of cartilage as a part of an osteochondral graft to replace focal regions of damaged articular cartilage. The plugs normally have a diameter of 5 to 15mm and can be used to fill defects of up to 9 cm² of superficial area. The graft technique is subdivided in autografts, when the graft is taken from an “unloaded” region of an articular surface of the same patient, allografts when the graft is taken from another patient (same specie), and xenografts, when the graft is taken from articular surfaces of another organic entity (animals). Each technique is selected in dependence of the bone quality, age of the patient and size of the defect. The principal advantages of using grafts are to provide a fully formed articular cartilage matrix and the potential for transplanting viable chondrocytes that can maintain the matrix.

Fibrous cartilage is appropriate to support and to transmit tensile loads, but in fact, in the majority of the cases, compressive loads act on synovial joints (e.g. at the knee), which could be optimally supported by hyaline cartilage. These joints are known as weight-bearing joints. For this reason cartilage transplantations from the same patient are currently used to fill the defect using autografts taken from “unloaded” regions at the joint. Since only a limited number of these regions are available, and such a transplantation in some cases represents a risk for the patient, allografts or predesigned biomaterials with a porous structure are frequently used to fill the defect (Evans, et al., 2004;Hunziker, 1999;Jackson, et al., 2001;Porter, et al., 2004). This practice is specially employed for the treatment of osteochondral defects in young patients (Buckwalter, 1995;Roughley, 2001) or in patients with defects with a diameter smaller or equal to 20 mm (Buckwalter, 1999;Ghadially and Ghadially, 1975;Hunziker, 2002;Kaar, et al., 1998;Newman, 1998;O'Driscoll, 1998;O'Driscoll and Fitzsimmons, 2001). In critical situations the usage of transplantations of stem cells from the bone marrow or from perichondrium are recommended (Barry and Murphy, 2004;Breinan, et al., 1997;Grande, et al., 2003;Meinel, et al., 2004;Oreffo and Triffitt, 1999;Rutherford, et al., 2003;Shapiro, et al., 1993;Wakitani, et al., 1994).

The first clinical reports about the usage of autografts showed a newly formed cartilage with 70-80% of hyaline consistence. However, the principal reason to use grafts in a long-term treatment is to conserve the original structure and the mechanical quality of the subchondral bone and within its capacity to transmit the acting external forces. In the present work, the effect of predesigned biomaterials on healing was evaluated using a tissue differentiation model to simulate osteochondral repair. Other treatments include the usage of hormones, medicaments or growth factors to promote or to improve osteochondral repair.

In order to calculate these parameters (elastic deformation, elastic modulus of Young, etc.) cartilage samples were tested under compressive loads (Appleyard, et al., 2003;Boschetti, et al., 2004;Franz, et al., 2001;Wei and Messner, 1998). With the use of mechanical tests, it is intended to reproduce the mechanical boundary conditions activated during joint movements (axial, bending and torsional loads). Three different methods are frequently used to estimate the cartilage material properties: confined compression, unconfined compression and indentation, whose differences are based on the boundary conditions assumed during testing (Bae, et al., 2004;Goldsmith, et al., 1996;Korhonen, et al., 2002;Korhonen, et al., 2002;Lyyra, et al., 1999;Ming, et al., 1997). In correspondence to these boundary constraints, changes in the fluid direction and thereby in the material properties are expected. When a joint is in contact with any another joint, e.g. at the knee joint, and the fluids do not have perpendicular restrictions relative to the direction of the compressive active load, these boundary conditions are referred to as confined compression. In contrast, when the fluids inside the cartilage do not have movement restrictions in the radial direction and the fluids could move only in a perpendicular direction of the acting load, an unconfined compression constraint is defined. An indentation test measures the deformation at the cartilage surface after compressive load application using an appropriated device (indenter; see below).

1. In 1995 Lyyra and co-workers presented a static indenter that was arthroscopically controlled (Lyyra, et al., 1995). The accuracy of the device was tested by comparison between arthroscopic measurements of the cartilage deformation and measurements realized with a conventional indenter from stress-relaxation tests in cadaveric knee joints. A good agreement between both indentation methods was thereby achieved. The authors conclude that such a device can be used in clinical practice.

2. A handheld dynamic indentation system was introduced by (Appleyard, et al., 2001). This novel device was developed to determine the dynamical mechanical properties of the articular cartilage in vitro. The manipulability of the device allows measuring the mechanical properties of all zones of an articular surface eliminating the accuracy problem of measurements in a joint with strong local differences in curvature. A map of forces acting on the articular surfaces can in this way be determined. With this device the dynamic shear modulus was calculated using the Hayes model.

1. Ultrasound elastomicroscopy imaging of soft tissues. This method consists of a compressive device to apply loads on a soft tissue specimen in vitro. During the test elastic radio frequencies are acquired from a scan, which produces formatted images of tissue response during load application. The compressive strain could then be calculated from the shape of the deformed specimens measured in the scanned images (Zheng, et al., 2004).

2. Using a high-frequency pulse echo ultrasound device, a Japanese research group (Kuroki, et al., 2004) determined the mechanical tissue quality during healing of osteochondral defects treated with autologous graft transplantation in rabbits. Kuroki et al. demonstrated that such a device indeed was not able to show any differences between the joint with the implant and the contra-lateral control joint at 0 days, as expected. The differences in the ultrasound signal intensity were detectable after 2, 4, 8, 12 and 24 weeks. Histological and microscopic analyses were additionally performed in which degradation, from 2 week, and detachment of the implant, between 12 and 24 postoperatively, were observed. A proper evaluation confirmed a close relation between the mechanical properties of the tissues (ultrasound signal intensity) and the degradation process. The intensity was reduced when damage was increased. Between week 2 and week 12, postoperatively, the intensity in the grafted joint was slower than that for the intact contra-lateral joint. Ultrasound can be applied in both in vivo and in vitro.

3. Articular damage as produced by rheumatoid arthritis (RA) could be detected by alterations in the fluid flow patterns inside the cartilage or by changes in the synovium. Synovial vascularization and fluid velocities can be detected by the Doppler technique. Ozgocmen and co-workers used the Doppler effect to determine flow patterns in joints with RA. Flow patterns have a strong correlation to intra articular bone and cartilage degeneration (Ozgocmen, et al., 2004). Power Doppler ultrasonography (PDUS) appears to be appropriate for determination of intraarticular vascularization and flow patterns of normal and osteoarthritic cartilage. Strunk et al. employed the Doppler technique for the first time (Strunk, et al., 2004).

4. Saarakkala and his group used an ultrasound indentation instrument to determine mechano-acoustic properties of the cartilage in vivo. After measurements of the speed of sound in the cartilage and its reflection coefficient, they succeeded in obtaining the elastic and dynamic modulus of the cartilage. These ultrasonic and mechanical parameters were compared for intact and degenerated cartilage specimens. A correlation between the ultrasonic responses with the grade of articular damage (estimated from the elastic modulus of Young) was found. The sound speed and the reflection coefficient increase for tissues with high values of dynamic and Young’s modulus. The authors developed a method to diagnose osteoarthroses in situ (Saarakkala, et al., 2004;Saarakkala, et al., 2004).

5. The mechanical environment of chondrocytes has been studied. Using different experimental techniques in combination with numerical methods the mechanical properties of chondrocytes were determined. Athanasiou and his group have developed a novel system that allows for the first time to realize creep indentation (Koay, et al., 2003) and unconfined compression tests (Leipzig and Athanasiou, 2005) on simple chondrocyte cells. The system configuration consisting of a force transducer cantilever, a sensor (laser micrometer), a motor and a PC running with LabView (software package to control load application) was able to apply constant compressive stress to individual adherent cells. After determination of the elastic modulus of Young numerical models were used to determine the other mechanical parameters (Poisson’s ratio, Aggregate modulus, permeability and viscosity). Previous approaches of measurements on cells have already been reported. Elastic modulus of Young on chondrocytes from normal and osteoarthritic cartilage samples were determined by applying pressure via cell aspersion (Jones, et al., 1999). Similarly, Poisson’s ratio of chondrocytes from normal and osteoarthritic cartilage were estimated applying pressure to the chondrocyte surface by micropipete aspiration (Trickey, et al., ). Using theoretical models, fluid pressure and shear strain state in chondrocytes have been calculated (Wu, et al., 1999). Employing finite element simulations, biphasic material properties of chondrocytes have been determined (Guilak and Mow, 2000). Other approaches use a combination of 3D confocal microscopy to determine volumetric changes in situ (deformation) during load application, and computer simulations to calculate additional mechanical properties (elastic modulus of Young, Poisson’s ratio, etc.) (Guilak, 2000). Another new optical technique is video microscopy, which has been used to measure radial deformations of human cartilage during classical tests of confined and unconfined compression. As a result, Poisson’s ratio, aggregate modulus and elastic modulus of Young can be determined. After evaluation of cartilage porosity, numerical approaches allow subsequent calculation of the related fluid parameters (permeability) (Boschetti, et al., 2004).

Generally ultrasonography appears to be a technique frequently used over the last years to estimate joint degeneration or to measure mechanical properties, and it has been employed to determine normal physiological values of diverse musculoskeletal regions. By determining physiological values in healthy tissues, the attempt was made to standardize musculoskeletal ultrasound measurements and to avoid erroneous interpretation of normal echoes as pathological values. Ultrasonography values from muscles, cartilage and tendons were measured and averaged in healthy individuals and presented by Schmidt (Schmidt, et al., 2004). Damage of articular surfaces can be determined with ultrasonography as well (Nieminen, et al., 2004).

Low-intensity ultrasound has been employed to stimulate differentiation of bone tissues (Korstjens, et al., 2004;Sakurakichi, et al., 2004;Tis, et al., 2002), of human mesenchymal stem cell into chondrocytes (Ebisawa, et al., 2004) or to analyze its viability proliferation and matrix production (Zhang, et al., 2003). Other authors have demonstrated that low-intensity ultrasound can even influence cartilage repair (Cook, et al., 2001;Zhang, et al., 2002) and cultured chondrocytes in 3D arrays (Nishikori, et al., 2002). Ultrasound has been employed to analyze the state of repair after graft transplantation (Hjertquist and Lemperg, 1971) or to predict histological findings in regenerated cartilage (Hattori, et al., 2004).

In order to attenuate and to distribute uniformly the acting external loads, during joint functionality the fluids are moving physiologically with a slow and continuous velocity through the cartilage (Buckwalter and Mankin, 1998;Buckwalter and Mankin, 1998). Fluid velocity is directly related with the mechanical material properties of the cartilage by the magnitude of its permeability. Permeability was initially calculated experimentally for soils using the Darcy’s law (Equation 1.1, Equation 1.2). With the Darcy’s experiment the hydraulic conductivity (K) of a material can be calculated. Permeability is a parameter that depends on hydraulic conductivity (Equation 1.2). Since the Darcy experiment implies some technical complications to measure this cartilage parameter, it is frequently calculated in an indirect way using the finite element method. As input data the elastic modulus of Young, Poisson’s ratio, compressive strains, and/or aggregate modulus of the cartilage are required.

Flow rate

(Equation 1.1)

Where

q =

ΔV₀/Δt flow rate (Rate of volume V in a time t)

=

Coefficient of permeability or hydraulic conductivity

A =

gross cross sectional area of flow

h =

total head (L)

L =

length of flow path

With the equation 1.1 (Darcy’s hydraulic conductivity) can be calculated.

Permeability

(Equation 1.2)

Knowing the permeability can be determined by equation 1.2

Where n =

fluid viscosity

= fluid density

g = gravitational acceleration

According to Forchheimer’s law, high flow velocities could reduce the effective permeability and consequently “chocking” pore fluid flow. As the fluid flow reduces, Forchheimer’s law approximates Darcy’s law (= 0).

Forchheimer’s law (permeability)
(Equation 1.3)

Permeability depending of porosity (n)
(Equation 1.4)

Void ratio (e) dependent on the porosity (n)
(Equation 1.5)

Porosity (n) depends on the strain (ε). Where n₀ is the initial porosity (Equation 1.6)

Where, =

is the volumetric flow rate of wetting liquid per unit area of the porous medium (the effective velocity of the wetting liquid)

=

is the fluid saturation. = 1 for a fully saturated medium, s = 0 for a completely dry medium.

n =

is the porosity of the porous medium.

=

is the fluid velocity

e =(dV_g +dV_t)

is the void ratio

dV_w

is the wetting fluid volume in the medium

dV_u

is the void volume in the medium

dV_g

is the volume of grains of solid material in the medium

dV_t

is the volume of trapped wetting liquid in the medium

(e)

is a “velocity coefficient”, which generally depends on the void ratio of the material

is the dependence of permeability on saturation of the wetting liquid such that = 1.0 at s = 1.0

g

is the density of the fluid

is the specific weight of the wetting liquid

k(e,θ)

is the permeability of the fully saturated medium, which can dependent on the void ratio (e) and/or temperature θ

u_w

is the wetting liquid pore pressure

x

is the position

In Abaqus, a finite element solver, the permeability is defined as:

(Equation 1.7)

where k

is the fully saturated permeability,

so that the Forchheimer’s law can be written as:

(Equation 1.8)

Remark: then have units of LT ^-1 . However, some authors use the hydraulic conductivity, which has units of L ² (or Darcy). To use the saturated permeability in Abaqus, it needs to be multiplied by / . Where is the kinematic viscosity of the wetting liquid (the ratio of the liquid’s viscosity to its mass density).

(Equation 1.9)

Permeability (k) can be determined by numerical approximations using the finite element method. With the equations 1.4, 1.5 and 1.6 the problem is completely defined. Experimentally the elastic modulus of Young (E), the Poisson’s ratio ( ) and the compressive strains ( ) are determined from an unconfined test. Assuming porosity, void ratio is calculated from the equation 1.5 (Abaqus uses void ratio as input data). By inserting n from the equation 1.6 in the equation 1.4 an expression of permeability depending on strains can be determined.

The permeability is selected such that the strains measured in the unconfined test experiment are reproduced. When the equation is satisfied, the permeability of the tissue has been found. Once k is known, with the values obtained for fluid velocity ( ) and pore pressure ( ), the fluid mechanical parameters of the tissue are fully determined (Equations 1.7 and 1.8).

Bones are able to adapt their internal structure in presence of mechanical stimuli generated after load application (Roux’s law). Different pathways for cell mechanotransduction have been identified to explain how this process could take place. The interpretation of these signals allows the development of tissue differentiation models, in which the mechanical cell environment is simulated and analyzed. For interpretation of this remodeling process Huiskes used the existence of an equilibrium state, originally proposed by Frost, in which cellular activity does not occur. An intact situation stays in the equilibrium state. Hence, equilibrium could be defined as a state in which the groups of cells responsible for growth or resorption are inactive and the balance between the acting external loads and the generated mechanical response causes the equilibrium. Remodeling is a complete cycle of cell activity in which after perturbations (prostheses, defects at bone, new boundary conditions) the equilibrium is newly reestablished. In this process different types of cells (osteoclasts and osteoblasts) are involved. This group of cells is called a basic multicellular unit (BMU). A BMU consists of about 10 osteoclasts and some hundreds osteoblasts. These groups of cells can be considered to be a mechanism of bone repair. Some authors believe that biochemical reactions are involved in the activation of these cells. Huiskes and his group propose that the BMU have some mechanosensors, which respond in the presence of external loads. Hence, deformations (a direct measurable consequence of the applied load) could be used as the mechanical stimulus to initiate a remodeling process. In other words, when an elastic, axial compressive load is applied to a tissue (for example bone), the derived compressive deformations act as mechanical stimuli to initiate remodeling.

Huiskes used this idea to develop a numerical model in which the knowledge of the equilibrium state of each bone region calculated in an intact bone situation was necessary. A stimulus for differentiation is activated as a reaction to the equilibrium alteration. The remodeling process stops when a tissue with the same mechanical properties as defined for an intact situation is completely reestablished in each affected region. The mandatory model of an intact bone must be calculated in order to know the strain magnitudes of the equilibrium state. After comparison of the straining of an affected and an intact situation, the necessary stimuli are calculated as a difference of these values. Huiskes schematized this process as a trilinear curve where maximal and minimal values of strains were used to define the range of change for each bone region. In this curve three points are defined to delimit the zones for growth, resorption, or equilibrium (dead zone) of each tissue as a function of a related stimulus (Huiskes, 1997;Huiskes, et al., 2000;Huiskes, et al., 1987) (Fig. 1.5). When the current strain falls below the value limiting the equilibrium zone, resorption sets in. In contrast, when the magnitude of the current strain is higher than the limits defined in the equilibrium zone, growth will occur. Following this schema, the elastic modulus of Young is continuously changed in an iterative process. With this model Huiskes calculated changes in the bone density (directly related to the elastic modulus of Young’s) that occur after implantation of a hip prostheses (Huiskes, 1990;Huiskes, 1997;Huiskes, et al., 1987;Kerner, et al., 1999;Mullender and Huiskes, 1995;Prendergast and Huiskes, 1995;Prendergast, et al., 1997;Vena, et al., 2000). This model was also applied to analyze the effect of stress shielding and how bone remodeling occurs after total hip replacement. In that study a special emphasis of the interface bone-implant was made.

Fig. 1.5: Huiskes’s model. Cell activity is represented by a tri-linear curve. High values of strain energy density conduce to growth activity, low values of strain energy density conduce to resorption activity. The dead zone represents cellular inactivity (Huiskes, et al., 1987).

Although Huiskes’s model showed a good correlation to the biological processes observed during and after a total hip implantation, this formulation was inappropriate to simulate fracture repair. Claes focused his work on the description of the mechanobiological process involved during fracture healing. After animal experimentation the total process from callus formation to cortical regeneration, the last state in a fracture repair, was documented and reconstructed through histological analysis at fixed points in time. In the Claes’s theory cells response to local strains and stresses, whose magnitude, position, and type could be determined. Claes modeled the geometry of the histological sections at the previously defined fixed points in time from his animal experimentation. Viscoelastic material properties taken from literature were used in these models. The finite element method was employed to determine mechanical strains and hydrostatic pressures for each tissue type (Fig. 1.6a).

Fig. 1.6a:Claes’s model. Characteristic limit values of strain and stresses are defined for each tissue type to simulate diaphyseal fracture healing (cortical remodeling through enchondral ossification) (Claes and Heigele, 1999).

The histological section showed how bone tissue was newly formed from an existent bone or cartilage region in an intramembranous ossification process. Cell differentiation for each tissue occurred in dependence of the acting strains or stresses whose magnitudes were previously determined with the computer model. Claes found that connective or fibrous tissue could be formed when hydrostatic pressures are higher than 0.15MPa and when the compressive strains are higher than 5% (Claes, et al., 1998;Claes, et al., 2002;Claes and Heigele, 1999).

Recently Claes’s group extended their model incorporating fuzzy logic

Fuzzy logic: Is a superset of boolean logic dealing with the concept of partial truth. Whereas classical logic holds that everything can be expressed in binary terms (1:yes; 0:no), fuzzy logic replace these simple boolean truth values with degrees of truth. This allows for values between 0 and 1, the concept of “maybe”. It allows partial membership in a set (i.e. elements can belong to different sets simultaneously). The method can use logical connectors to describe unclear events or events that cannot be represented with unique true values. A membership function is normally required to adjudicate the grade of certainty of the variable to be evaluated. In this form unclear events can be treated with IF/THEN rules for final association of a deterministic success.

Fig. 1.6b: Fuzzy logic in combination with mechanical stimulus applied to simulate healing of trabecular fractures (intramembranous ossification). Fixed rules including biological factors regulate building or resorption of different tissue types (Shefelbine, et al., ).

Claes’s group demonstrated the possibility of using a general tissue differentiation model for fracture healing. In this form, the fuzzy logic model, which was initially employed to simulate diaphyseal fracture healing (Ament and Hofer, 2000), was further developed and applied to simulate metaphyseal fracture healing and bone remodeling. The first one refers to healing when the fracture occurs in the cortical bone. Cortical bone fractures heal by enchondral ossification: after a fracture, a callus replaces the formed haematoma. The callus is differentiated to fibrous and later to hyaline cartilage before it is replaced by cortical bone. The second one, metaphyseal fractures, compromises the trabecular bone. Trabecular bone fractures heal through intramembranous ossification: The bone is remodeled by activation of BMUs (basic multicellular units); osteoclasts and osteoblasts absorb and build new bone directly in dependence of biological and mechanical boundary conditions without cartilage formation. As a mechanical stimulus for differentiation they used octahedral strains in combination with hydrostatic strains.

Fuzzy logic is however a controversially discussed method because it anticipates the response of the healing process (or the system to be evaluated ) showing logically a high dependency on the fixed rules. However, the major contribution of Claes’s proposal is its intent to unify the fracture healing process for cortical and trabecular bone, allowing the analysis of fracture healing for different types of bones and different fracture localizations. Perhaps even more important is the point that his model allows the incorporation of biological factors observed during in vivo fracture healing such as different rates of vascularization and concentration of growth factors for the early and the late stages of healing.

Prendergast analyzed fracture repair under another perspective. Since following haematoma formation in the first stages of fracture repair only connective tissue is recognizable, which has a higher content of water, Prendergast proposed that cells react principally under the effect of this fluid phase. Fluid velocity and strains as a result of external loads should play an important role in the development of appropriate mechanical boundary conditions to initiate the repair process. Animal experimentation was used to define strain limits to start and to maintain cellular differentiation. During healing it was observed that micromovements at the bone surface stimulate bone repair. Using a biphasic, pore elastic model the mechanical shear strains and fluid velocities for each differentiated tissue during the repair process can be calculated. A mechanoregulation schema that combines these two mechanical parameters was then proposed to describe fracture healing (Fig. 1.7).

Fig. 1.7:Prendergast’s model. Prendergast developed a mechanoregulation schema regulated by shear strains and fluid velocities defined for each tissue type (Lacroix and Prendergast, 2002).

It was determined for example that a combination of high values of shear strains and fluid velocities cause distortion in the mesenchymal stem cells acting as stimulus to differentiate these cells to connective tissue. In a similar way cartilage is formed after a combination of low values of shear strains and slow fluid velocities (Lacroix and Prendergast, 2002;Prendergast, 1997;Prendergast, et al., 1997;Prendergast, et al., 1996).

Bailon-Plaza developed a predictive fracture healing hyperelastic model, in which growth factors were taken into account as an important parameter during repair. Plaza and co-workers proposed that growth factors are responsible for initiation and regulation of biological process of fracture healing. From in vivo studies the chondrogenic and osteogenic effect of two different growth factors were determined. The ability of these growth factors to regulate differentiation appears to be strongly related with the number of cells. A numerical implementation of this theory was used to simulate fracture repair. The model was able to predict and to demonstrate that the formation of osteogenic growth factors from osteoblastic cells and the duration of their production are necessary to obtain a full bone restoration after trauma. In her computer model dilatational and deviatory strains were used as stimuli to regulate the production of growth factors present principally in the first stages of fracture repair. The material properties used in her model were taken from literature (Bailon-Plaza, et al., 1999;Bailon-Plaza and van der Meulen, 2001).

The work of Lill and co-workers is briefly introduced. In the biomechanics laboratory of the Musculoskeletal Research Center in Berlin, a previous study considering different bone qualities was realized (Lill, et al., 2002). Some parameters of humeral human bones were determined: stiffness, density distribution, trabecular bone orientation and volume bone fraction. Both intact and fractured human bones, stabilized with different medical devices, were mechanically tested and analyzed. The goal of Lill and his group was to determine favorable implant stiffness for fracture healing, taking into account the effect of the bone density distribution. Important implications for the treatment of fractures in osteoporotic patients were established.

Histomorphometrical, CT and MRI analysis were performed on 24 freshly harvested human cadaveric humeri. Histomorphometric analysis evaluated structural parameters (tissue volume to bone volume ratio, trabecular thickness), connectivity (number of nodes

Nodes: trabecular intersections at the cancellous bone.

The center of the trabecular structure connects the center of the gleaned cavity. A correlation between the structural parameters of the trabecular network, its localization and the estimated bone quality was determined: higher bone stiffness was found in regions of high trabecular density and elevated number of nodes compared with regions of low trabecular density. The project of Lill and his group aimed to obtain knowledge of distribution, microstructure, and quality of bone in the humeral head, which allows the remaining bone stock to be used effectively, even in elderly patients, with a minimally invasive approach and maximum mechanical stability (Hepp, et al., 2003;Lill, et al., 2002;Lill, et al., 2001;Lill, et al., 2003). According to Lill’s findings, bone quality appeared to be an important parameter to be evaluated. Therefore, the effect of the bone quality and its relation to the mechanical conditions (physiological loads) in intact and fractured bones during bone healing was then evaluated in this project.

In a subsequent study, 35 fresh human humeri were used to perform mechanical testing in order to determine in vitro characteristics such as stiffness of different conventional and new devices used to stabilize proximal humeri fractures. The implants tested under static and cycling loading included humerus T-Plate (HTP), the cross-screw osteosynthesis (CSO), the unreamed proximal humerus nail with spiral blade (UHN), the Synclaw Proximal Humerus Nail (Synclaw PHN) and the angle-stable Locking Compression Plate Proximal Humerus (LCP-PH). Three clinical load cases were evaluated: axial compression, torsion and varus bending. The results showed that the evaluated conventional devices (HPT, CSO, UHN, Synclaw PHN) presents a higher stiffness under static load than that encountered in the LCP-PH. Additionally, the torsional stiffness was strongly reduced with exception of the LCP-PH.

Lill and co-workers concluded that implants with low stiffness and corresponding elastic properties could minimize peak stresses, which could be related to the early loosening and failure of the interface implant-bone. Therefore low stiffness implants seem particularly suitable for fracture fixation in osteoporotic bones (Lill, et al., 2003).

The animal experiment realized by Bail and co-workers is briefly summarized (Bail, et al., 2003). Osteochondral defects of 6 mm diameter and 1.5 mm depth from the subchondral bone plate were created at the left femoral condyle in 18 Yucatan minipigs.

Under general anesthesia an osteochondral defect was created at the lateral surface of the trochelar groove of the left hind limb. The osteochondral defect was 6 mm in diameter and 1.5 mm in depth from the osteochondral junction (Fig. 2.2). To minimize soft tissue damage a sharp tube was pressed into the joint cartilage to the osteochondral border for guiding a 6 mm drill. A sleeve on the drill ensured the depth of the osteochondral defect. All wounds were sutured and covered with spray bandage. Animals received flunixin (Finadyne, Essex, Great Britain) as an analgesic for the first 7 days. Six animals were sacrificed after 4 weeks, nine after 6 weeks and the remainder after 12 weeks (Bail, et al., 2003).

In order to achieve a full load condition at the affected joint, an osteotomy at the contra lateral hind leg stabilized with a 8 to 10 hole DPC was created after surgery. To obtain a continuous load condition the animals were allowed to move freely after defect creation. The animals were on average 17.1 months old (10.5 to 30 months).

The following boundary conditions were reproduced in the model definition to perform osteochondral healing simulation: joint geometry and the defect size were modeled to resemble the mini-pig’s knee joint, and the total load was continuously applied from the initial defect situation for the rest of the healing process miming the in vivo load situation.

The animals walked on a compressive sensible platform. The measurements were made one day before surgery, three days after surgery and each week until the sacrifice of all animals. A complete gait cycle was defined as the enfolded step realized for a hind leg simultaneously with a foreleg. For each member 7 imprints were recorded (to minimize errors in the measurements). The weight of each animal was controlled during the realization of each gait analysis. After gait analyses an equivalent compressive load of 1.35 MPa was calculated from reaction forces of 1 body weight (BW) acting on an average area of 150 mm² at the femoral condyle.

The histological sections were stained with Safranin-Orange van Kossa and Safranin-Light Green. This staining allows a clear identification of the formed tissue types. Safranin-Orange van Kossa stains the calcified tissues (e.g. subchondral bone) black and the non-calcified tissues (hyaline cartilage, fibrous cartilage and connective tissue) orange. Safranin-Light Green stains connective tissue and bone green, cartilage and cell nucleus are colored red. The mechanical tissue quality of the total group was slower than the corresponding of an intact situation.

The new formed tissues (hyaline and fibrous cartilage), the remodeled bone, and the remaining defect were quantified during healing by a histomorphometrical analysis (KS400 image analysis system, Zeiss, Germany). In three regions of interest (ROI), localized within the subchondral bone at the walls and at the basis of the defect, the structural orientation and the trabecular bone fraction were determined. For control purposes all histological and histomorphometrical analysis were made at the same regions of the contra lateral uninjured femur condyle. All statistical analyses (p<0.05) were carried out using a commercially available package (Statistical Package for Social Sciences, SPSS Inc., Chicago, USA) (Bail, et al., 2003). The trabecular bone fraction was increased indicating active remodeling regions.

Fraction of cancellous bone increased from 25% of the total area to 29.1 ± 10.6% at 4 weeks, subsequently to 33.0 ± 13.9 % at 6 weeks and to 33.2 ± 8.7 % at 12 weeks. Hyaline cartilage was increased from initially 0% to 8.1 ± 6.7 % at 4 weeks, 17.0 ± 12.5 % at 6 weeks, and 33.1 ± 25.4 % at 12 weeks. The fibrous tissue filling the defect decreased from 47.7 ± 14.2 % at 4 weeks to 35.2 ± 13.2 % at 6 weeks, and to 24.2 ± 18.7 % at 12 weeks. The defect region was reduced from initially 75% of the total area to 15.2 ± 7.71 % at 4 weeks, to 14.8 ± 9.0 % at 6 weeks, and to 7.1 ± 2.9 % at 12 weeks. During healing it was found that the trabecular bone fraction increases indicating active remodeling regions (Bail, et al., 2003).

Additionally, Bail et al. reported a complete macroscopic and microscopic description of the repair process. New techniques such as a reproducible immunohistochemical color protocol for cartilage and bone repair analyses were developed.

1. The implant LCP-PH was selected for analysis. After comparison between different implants to stabilize the proximal fractures, LCP–PH showed flexibility without compromising fracture fixation.

2. Two bones with known DEXA distribution, which were stabilized with a LCP-PH, were selected for reconstruction and analysis.

3. In this thesis the analysis, results and conclusions related to bone-joint mechanics, which was studied in humerus specimens under physiological loads, were identified as and reported under the name “humerus project”.

From Bail’s animal experiment the following data was used:

1. The histological sections at 4, 6 and 12 weeks were employed. In this work his data was used first to reconstruct the geometry of the histological sections and to analyze these models numerically in order to determine straining of the tissues during healing and factors for growth and resorption for each tissue type. The stained tissues were used to identify each tissue (Fig. 3.5). Secondly, the spontaneous repair process observed in the histology was used to compare qualitatively the in vivo healing outcome to the simulated healing.

2. The histomorphometrical data at 4, 6 and 12 weeks was employed to validate the model quantitatively as explained in the material and methods section.

3. From the gait analysis the reaction forces were calculated. These measurements were employed to estimate the in vivo load condition on the joint and applied as mechanical boundary condition during the simulated healing.

The study of spontaneous repair of osteochondral defects using animals for experimentation were realized in line with a doctoral thesis in veterinary medicine entitled “Histologische, immunhistologische und histomorphometrische Untersuchungen der Wirkung von systemisch applizierten Wachtumshormon auf einen osteochondralen Knorpeldefekt am Yucatan-Minischwein” (histological, immunohistological and histomorphometrical analysis of the effect of systematical applied growth factors on an osteochondral defect in Yucatan minipigs) (Klein, 2001).

In this thesis all sections related to osteochondral defect healing have the title “Galileo project”. Historically, Galileo Galilei is considered to be a pioneer in the field of biomechanics. He not only used the term “mechanics” for the first time to relate acting forces, systems and its response, but also observed and applied such concepts to biological organisms. Galileo supposed for the first time a relation between bone and its mechanical boundary conditions. His contribution to biomechanics was the basis for further development of bone remodeling and tissue differentiation (a more general formulation). Such concepts have been refined, further developed and are recently called mechanobiology (van der Meulen and Huiskes, 2002). Therefore, the name “Galileo” was used to identify the study of the mechanical aspects of the biological process involved in osteochondral healing.