Biomechanics of integrative cartilage repair

Osteoarthritis and Cartilage (1999) 7, 29–40  1999 OsteoArthritis Research Society International Article No. joca.1998.0160, available online at http://www.idealibrary.com on

1063–4584/99/010029+12 $12.00/0

Biomechanics of integrative cartilage repair

B T A  R L. S Department of Bioengineering and Institute for Biomedical Engineering, University of California, San Diego Summary Cartilage repair is required in a number of orthopaedic conditions and rheumatic diseases. From a macroscopic viewpoint, the complete repair of an articular cartilage defect requires integration of opposing cartilage surfaces or the integration of repair tissue with the surrounding host cartilage. However, integrative cartilage repair does not occur readily or predictably in vivo. Consideration of the ‘integrative cartilage repair process’, at least in the relatively early stages, as the formation of a adhesive suggests several biomechanical approaches for characterizing the properties of the repair tissue. Both strength of materials and fracture mechanics approaches for characterizing adhesives have recently been applied to the study of integrative cartilage repair. Experimental configurations, such as the single-lap adhesive test, have been adapted to determine the strength of the biological repair that occurs between sections of bovine cartilage during explant culture, as well as the strength of adhesive materials that are applied to opposing cartilage surfaces. A variety of fracture mechanics test procedures, such as the (modified) single edge notch, ‘T’ peel, dynamic shear, and trouser tear tests, have been used to assess Mode I, II, and III fracture toughness values of normal articular cartilage and, in some cases, cartilaginous tissue undergoing integrative repair. The relationships between adhesive biomechanical properties and underlying cellular and molecular processes during integrative cartilage repair remain to be elucidated. The determination of such relationships may allow the design of tissue engineering procedures to stimulate integrative cartilage repair. Key words: Biomechanics, Articular cartilage, Repair, Fracture mechanics.

Introduction

requirements for the functional and durable repair of cartilage defects. Cartilage lesions have been classified as those resulting from superficial lacerations, deep (fullthickness cartilage) injuries either penetrating or not penetrating the subchondral bone, or blunt impact [2, 5, 6]. From a macroscopic viewpoint, the complete repair of cartilage defects, in the most general scenario, requires two processes (Fig. 1): (a) the integration of the repair tissue with the surrounding host articular cartilage; and (b) the filling of the bulk of the cartilage defect with tissue that is characteristic of normal articular cartilage. These two processes can occur concurrently and may involve similar cellular and molecular mechanisms. This review focuses on the ‘integrative repair process’ (see below), which appears particularly problematic, and in particular, the biomechanical aspects of integrative cartilage repair. Integrative cartilage repair is required in a number of orthopaedic conditions and rheumatic diseases. Fibrillation, erosion, and vertical and horizontal cracking of the cartilage (Fig. 1A) are histopathological hallmarks of the osteoarthritic

CENTURIES ago, Hunter commented that ‘‘from Hippocrates to the present age it is universally allowed that ulcerated cartilage is a troublesome thing and that, once destroyed, it is not repaired’’ [1]. During the last two decades, experimental studies of cartilage repair in animals have led to human clinical studies [2], and the commercialization of cellular therapies (Carticel, Genzyme Tissue Repair, Cambridge, MA [3]) and specialized instruments (Acufex MosaicPlasty, Smith and Nephew, Andover, MA; COR System, Innovasive Devices, Inc., Marlborough, MA; OATS, Arthrex, Inc., Naples, FL) for the repair of isolated cartilage defects. While these innovations have spurred much interest, a number of important biologic and biomechanical issues remain to be resolved [4]. Relatively few studies have assessed the biomechanical quality of repair tissue or the

Corresponding author: Dr Robert L. Sah, Department of Bioengineering, 9500 Gilman Dr., Mail Code 0412, University of California-San Diego, La Jolla, CA 92093-0412. Tel.: (619) 5340821; Fax: (619) 534-6896; E-mail: [email protected]

29

30

T. Ahsan and R. L. Sah: Biomechanics of integrative cartilage repair

FIG. 1. Schematic of cartilage defects, intrinsic cartilage repair, and putative therapeutic interventions. A: Fissure. B: Chondral defect. C: Osteochondral defect.

disease [7]. Blunt trauma can also generate fissures, which can be oriented vertically, horizontally, or obliquely [8–10]. In chondral and osteochondral fracture, as well as osteochondritis dissecans, the reduction of the fracture or osteochondral fragment leads to the direct apposition of cartilage to cartilage. In all of these pathological conditions, integrative repair between opposing cartilage surfaces is required. In surgical [5] and tissue engineering [11] procedures to replace damaged articular surfaces with grafts of osteochondral, chondral, or cell-laden synthetic tissues (Fig. 1B,C), integrative repair of the graft tissue with host cartilage is also required. Unfortunately, integrative repair of cartilage does not occur readily or predictably in vivo. Although the fate of individual cartilage fissures in osteoarthritis is di$cult to assess, the natural history of the disease suggests that such fissures do not heal spontaneously. Indeed, it is generally accepted that cartilage defects, left untreated, lead to progressive degeneration of the surrounding and opposing articular cartilage. In experimental animal studies of the natural history of ‘scarification’ injury to the articular cartilage (i.e., multiple cuts extending to but not through the zone of calcified cartilage) of a normal joint, light and electron microscopic analysis have revealed that

such lesions are generally stable over a period of up to two years, with neither integrative repair nor a marked or consistent degeneration [12–17]. In studies of the natural repair following a penetrating defect, histological analysis indicates a relatively vigorous cellular response. Nevertheless, the repair tissue often fails to integrate with the surrounding cartilage [18, 19]. The repair tissue forming after the transplantation of periosteal and perichondrial grafts [20, 21] also does not adhere firmly and consistently to the surrounding cartilage. It remains to be determined if the transplantation of exogenous cells, either chondrocytes or chondroprogenitor cells [2, 22–24], or the stimulation of cellular processes with growth factors [25] augments the integrative repair process. The mechanical environment resulting from joint loading may a#ect the initiation, progression, and repair of cartilage fissures, as well as integrative repair processes in other scenarios. Mechanical factors have long been implicated in the regulation of cartilage metabolism in vivo [26]. Joint loading induces biophysical phenomena that may directly a#ect the cartilage extracellular matrix or indirectly a#ect the matrix by regulating the metabolism of the endogenous chondrocytes [27]. While a number of theoretical studies have

31

Osteoarthritis and Cartilage Vol. 7 No. 1

FIG. 2. E#ects of joint loading on cartilage biomechanics and integrative repair loading in vivo. A: Integrative repair of a fissure in cartilage. Joint loading that can result in expansion of repair tissue in the tangential direction (arrows). B: Integrative repair between a region of repair and normal cartilage. Joint loading that can result in relatively large gradients (arrows) in tissue displacement and stress across the interface between repair tissue and normal cartilage.

assessed the biomechanics of the contact of normal joints using elastic [28, 29] or biphasic [30] models, relatively few studies have examined cartilage biomechanics in joints where defects exist and repair processes may be present. Two-dimensional biphasic finite element analysis of the perforation of subchondral bone has been predicted to lead to increased tensile and, in particular, shear stress in the cartilage near the perforation site [31]. Biphasic finite element analysis has also predicted the existence of deformation gradients across the interface between repair and normal cartilage [32]. Elastic models examining the propensity for an existing crack within articular cartilage to propagate during joint loading have found that, depending on the crack and joint geometry, tensile [33, 34] and shear [35] stresses develop at the crack interface. Although these theoretical studies involved simplified models of cartilage, they do provide insight into the probable existence of tensile and shear stresses that may counter integrative repair processes in vivo (Fig. 2).

Experimental and theoretical methods traditionally used to study the biomechanical properties of cartilage are di$cult to apply to repair tissue and the integrative repair process. The homogeneous material properties that are typically estimated from the results of traditional compressive or tensile tests only approximately describe the behavior of articular cartilage, which is known to be both inhomogeneous and anisotropic [36–38]. In repair situations where remodeling tissue is adjacent to relatively normal cartilage, tissue properties are likely to be particularly inhomogeneous. However, isolating su$cient quantities of repair tissue for independent testing is technically di$cult. Furthermore, it is not clear whether estimates of the material properties of repair tissue (e.g., equilibrium confined compressive modulus, Poisson’s ratio, or hydraulic permeability [38]) would reflect the biomechanical interaction between repair tissue and host cartilage, and thus the integrative repair process. In the integrative cartilage repair process, it is reasonable to postulate that repair tissue forms a molecular bridge between the opposing cartilaginous tissue surfaces. In this respect, repair tissue functions as an adhesive, a material at ‘‘the surfaces of two solids [that] can join them together such that they resist separation’’ [39]. Consideration of integrative cartilage repair, at least in the relatively early stages, as the formation of an adhesive material suggests several biomechanical approaches for characterizing the properties of the repair tissue. Both strength of materials and fracture mechanics approaches for characterizing adhesives and materials have recently been applied to the study of integrative cartilage repair. The failure properties of the adhesive material after integrative cartilage repair may be compared to the failure properties of normal articular cartilage. When adhesive materials are subjected to load, either adhesive or cohesive failure may occur. Adhesive failure is the rupture of an adhesive bond such that the separation is at the adhesive-adherend interface. In contrast, cohesive failure is the rupture of an adhesive bond such that the separation is within the adhesive [40]. In the experimental studies of cartilage repair that are reviewed below, it remains to be determined whether the mechanism of failure was primarily adhesive or cohesive. Nevertheless, in this paper, for convenience, the failure of repair or bonded cartilage will be termed adhesive failure. These adhesive properties may be compared to the failure (cohesive) properties of normal articular cartilage to assess the quality of the repair process.

T. Ahsan and R. L. Sah: Biomechanics of integrative cartilage repair

32

FIG. 3. Adhesive strength testing of integrative cartilage repair. A: Single-lap test of pairs of articular cartilage blocks and a representative result ([42], graph reproduced with permission). B: Test of adhesive strength between the cartilage of osteochondral cores [43].

Strength of materials One approach to characterizing the mechanical failure of a material consists of determining the normal or shear strength, defined as the ultimate stress that is sustained just before failure [41]. When traditional (non-biological) engineering materials are bonded together in a joint, the strength is the maximum stress at which the adhesive can still maintain the integrity of the joint [40]. Experimental configurations, traditionally used to test structural adhesives, have been adapted to determine the strength of the biological repair that occurs between sections of bovine cartilage during explant culture [42], as well as

adhesives that are applied between sections of cartilage [43, 44]. A modification of a standard structural mechanical test method for estimating the strength of adhesives in a single-lap configuration [45] was used by Reindel et al. [42] to determine the adhesive strength of the repair tissue (Fig. 3A), noted previously [46], to form between cartilage surfaces during in-vitro culture. Cartilage blocks were harvested from calves and adult steers. Pairs of cartilage blocks were maintained in partial apposition by the application of load to the overlap area, providing a normal stress of 0.1 kPa. After integrative repair during incubation in vitro, the adherent cartilage blocks were separated by applying a constant rate of uniaxial positive displacement, slow enough to minimize fluid pressurization during the test, and the resultant load was measured (Fig. 3A). The adhesive repair tissue was considered to have failed at the maximum load, and the adhesive strength was calculated as the measured ultimate load divided by the original overlap area. The average adhesive strength for calf and adult bovine tissue incubated for 3 weeks in culture medium supplemented with 20% fetal bovine serum were 20 kPa and 34 kPa, respectively (Table I). Since the tensile modulus and strength of normal bovine and human cartilage may be as high as four orders of magnitude greater than these values [38, 47, 48], the adhesive properties and geometry of the repair tissue, rather than the properties of the cartilage adherends, are likely the predominant factors influencing the measurement. Since the intrinsic integrative repair process of post-natal articular cartilage seems limited, the utility of using biochemical cross-linking agents to provide a relatively rapid and initial integration of apposing cartilaginous surfaces has been investigated. Tissucol (Immuno AG, Vienna, Austria) is a commercially available fibrin-based sealant that is prepared from pooled plasma of multiple donors and used for its tissue adherence and hemostatic potential [43]; the covalent stabilization of the

Table I Adhesive strength in integrative repair Tissue Articular Cartilage: Articular Cartilage: Articular Cartilage: Articular Cartilage: Articular Cartilage: Meniscus: human

calf bovine bovine bovine human

Treatment

Strength [kPa]

Reference

Explant culture Explant culture Tissue transglutaminase Fibrin sealant Photochemical welding Photochemical welding

20 34 25 29 120 180

[42] [42] [43] [43] [44] [44]

33

Osteoarthritis and Cartilage Vol. 7 No. 1

fibrin clot is mediated by plasma transglutaminase (activated Factor XIII) [49]. Tissue transglutaminase, a related enzyme, can also catalyze the formation of stable polymer networks [49]. Ju¨rgensen et al. [43] assessed the adhesive strength of Tissucol and tissue transglutaminase using osteochondral cylinders harvested from adult bovine shoulder joints. The cartilage was cut to form a planar surface, and the adhesive was applied to the cartilage surfaces. Then, one cylinder was placed upon another to create a cartilage– cartilage interface, and a normal stress of 66 kPa was applied to maintain tissue apposition. Samples were tested by measuring the lateral force required to dislodge the cylinders (Fig. 3B). Incubation with Tissucol or tissue transglutaminase resulted in a time-dependent increase in adhesive strength, computed here as the maximum force divided by the interface area. For an incubation time of 10 min at 37C in 50% humidity, Tissucol and tissue transglutaminase (1 mg/ml) resulted in an adhesive strength of 29 kPa and 25 kPa, respectively (Table I), significantly higher than that achieved by controls incubated with bu#er solution. Enzymatic pretreatment of tissue surfaces with chondroitinase AC before application of tissue transglutaminase resulted in a modest (30–60%) increase in the adhesive strength. Photochemical crosslinking methods have also been developed to provide an initial integration between apposing cartilaginous tissues. Using the same single-lap configuration described above [42], Jackson et al. [44] investigated the adhesion produced by photoactivation of 1,8-naphthalimide dye that was applied between articular cartilage explants. A 12 mM solution was applied to the opposing faces of tissue blocks, the blocks were compressed together between glass slides with a 294 kPa normal stress, and the solution was activated by irradiation at a wavelength of 458 nm and a power density of 200 mW/cm2. While the application of laser irradiation alone and dye solution alone did not promote adhesion, the laser activated photoactive dye resulted in adhesive strength of 120 kPa (Table I) for human articular cartilage. Similar experiments with human meniscus resulted in an adhesive strength of 180 kPa. The assessment of integrative cartilage repair by quantitating adhesive strength has advantages and disadvantages in comparison to the fracture mechanics methods described below. Implementation of the adhesive strength tests, with relatively large surface areas of repair tissue being tested, results in relatively large loads. Alternative configurations to measure adhesive strength involve the application of torsional forces, such as to

cylindrical or annular (‘napkin-ring’) samples that form a butt joint [50]. However, the actual stress distribution in each of these tests is complex and di$cult to determine. For example, in the singlelap configuration, the stress varies over the area of the adhesive even for ‘ideal’ adherends and adhesives [40], yielding a structural property that depends on the test geometry instead of an intrinsic biomechanical property. Moreover, for articular cartilage, such complex states of stress may be associated with significant interstitial fluid pressurization [38], thus complicating interpretation of the measured load response of the adherends. Thus, although the adhesive strengths achieved after several weeks of biological integrative repair are similar to those attained by application of Tissucol and tissue transglutaminase and less than those attained by photochemical welding (Table I), a direct comparison is di$cult due to the di#ering test geometries, di#ering stresses within the adhesives and adherends, and di#erent sources of adherend cartilage tissues. The methods of determining adhesive strength, described above, are particularly useful in early stages of integrative cartilage repair where the interface region is relatively weak. However, as integration at the interface improves and the function of the repair tissue approaches that of the adjacent cartilage, the stress distribution in the adhesive will become more complex and may result in the method becoming less suitable for assessing the adhesive strength of the repair tissue [45]. In addition, it is di$cult to compare the estimates of strength of the repair tissue to that of normal cartilage since di#erent tests are usually used. On the other hand, during the final stages of a successful cartilage repair process, it may be possible to use traditional techniques [38, 47] of determining tissue strength. It would be desirable to use the results from the in-vitro studies of adhesive strength to guide the study of therapeutic interventions in vivo. The in-vitro specimen geometries allowed assessment of integrative repair between sections of cartilage that were approximately parallel to and just below the articular surface. Such repair may be more relevant to fissures that extend horizontally than to those fissures that extend vertically, since the orientation of matrix components, such as collagen fibrils, adjacent to the fissure may have a significant e#ect on the repair process. Nevertheless, the cellular and molecular mechanisms involved in such an integrative repair process may be similar whether integrative repair is required to occur across a vertical or horizontal plane within articular cartilage.

34

T. Ahsan and R. L. Sah: Biomechanics of integrative cartilage repair

Fracture mechanics An alternative approach to describing the integrative repair process involves the application of the principles of fracture mechanics [51]. Fracture mechanics is the study of the failure of materials with the assumption that a material includes flaws. The propagation of these flaws to cause failure is dependent on the applied stress and the fracture toughness of the material. Since fracture mechanics was originally developed for the study of cohesive failure of materials and then extended to study the integrity of adhesive-bonded joints, it is natural to use this approach to compare the fracture toughness of normal and repaired cartilage. The two main subdivisions within the field of fracture mechanics utilize either a stress intensity criterion, where the critical stress intensity is a measure of fracture toughness, or an energy criterion, where fracture toughness is described by the critical energy release rate. The stress intensity approach relates the local stress at the crack tip to the applied stress and crack geometry. Although this approach can be adapted to characterize crack propagation with small plastic zones, the critical stress intensity is dependent on linear elastic material properties, and the obtained values should be treated with some caution [52]. The energy criterion approach has been adapted to characterize toughness for nonlinear elastic and viscoelastic-plastic materials [51]. The energy criterion of fracture mechanics states that crack extension occurs when the energy available for crack growth is su$cient to overcome the resistance of the material. The fracture toughness is described by the energy required to extend a crack per unit of crack extension [51, 53]. Such energy methods have been applied to cartilage and other soft biological tissues. In elastic materials, the fracture toughness is the work of fracture, a measure of the energy needed to create new free surfaces. However, in general and for the study of cartilage, the rate of work of external loads is dissipated not only by fracture, but also by elastic energy stored and viscous, frictional and plastic energy dissipation [54]. These other energy terms are dependent on specimen size, geometry, orientation and testing strain rate. When the contribution of the fracture term can be isolated, a material property describing fracture toughness is obtained. A variety of di#erent test configurations and procedures have been used to assess the fracture properties of articular cartilage and other connective tissues. In comparing the fracture toughness values of di#erent biological tissues, it is import-

FIG. 4. Schematic of fracture modes. A: Mode I, opening. B: Mode II, in-plane shear. C: Mode III, out-of-plane shear.

ant to consider the test configuration. The mode of loading at the crack tip usually a#ects the measured fracture toughness. Mode I loading opens a crack by inducing tensile stresses normal to the crack plane (Fig. 4A). Mode II loading propagates a crack between two surfaces by inducing in-plane shearing loads (Fig. 4B). Mode III loading extends a crack by transverse (out-ofplane) shearing (Fig. 4C). Even in isotropic, linear elastic materials, the di#erent modes of loading result in distinct values for fracture toughness [51]. It is di$cult to control the direction of crack propagation in fiber-reinforced composite materials [55], and the same di$culty arises in articular cartilage [56, 57]. Nevertheless, the fracture toughness values, estimated from energy release rates, for the di#erent loading modes in di#erent biological tissues have the potential to provide insight into the function of normal cartilage as well as the adequacy of tissue remodeling during integrative repair. Since tensile stresses may induce crack propagation in cartilage, the determination of Mode I fracture properties are of particular interest. The modified single edge notch test (MSEN) has been used to quantify the energy required during Mode I crack extension perpendicular to the articular surface (Fig. 5A) in canine cartilage using a 0.2 mm thick osteochondral slices from adult mongrel canine patella [54]. A cut was made through the subchondral bone to make an initial crack that

35

Osteoarthritis and Cartilage Vol. 7 No. 1 Table II Fracture energy of normal and healing biological tissues Mode I

II III

Tissue

Fracture energy [J/m2]

Reference(s)

Articular cartilage: canine Articular cartilage, repaired: bovine Synthesized cartilaginous tissue: rabbit Meniscus: rabbit Meniscus, repaired: rabbit Corneal stroma: human Corneal stroma: rabbit Corneal stroma, repaired: rabbit Skin stratum corneum: human Skin, grafted: rat Osteochondral junction: bovine Articular cartilage: canine Aorta: pig Skin: rat

140–1500 16 1000 1500* 290* 140 98 36 3600 20 36,000–58,000 240–1200 1800 25,000

[54] [61] [58] [62] [62] [64, 86] [65, 66] [65] [87] [67] [52] [54] [68] [69]

*Fracture energies calculated assuming a 1 mm meniscal thickness.

extended 0.2 mm into the deep layer of cartilage. To avoid aberrant crack propagation in the surface tangential zone [57], the superficial region of cartilage was removed. The specimens were gripped at the bone and separated at a relatively low rate (to minimize fluid flow-dependent e#ects) so that the crack propagated from the deep zone towards the articular surface of the cartilage. While fissures in osteoarthritis may propagate from the articular surface toward the bone, the use of the subchondral bone for gripping the samples and propagating the crack in the opposite direction is convenient experimentally. Finite element analysis was used to obtain a relation between the fracture toughness and experimentally measured parameters. Experimental studies, together with the theoretical analysis, resulted in estimates of the fracture toughness of normal canine cartilage that varied from 140 J/m2 in samples that fractured sharply to 1460 J/m2 in samples that exhibited a more blunt pattern of fracture (Table II). These values are similar to the fracture toughness, obtained using a single edge notch test, for tissue synthesized in vitro by rabbit chondrocytes after isolation and then 8 weeks of culture (300–1200 J/ m2 [58, 59], Lewis, personal communication). A modification of a standard fracture mechanics ‘T’ peel test method (Fig. 5B, [60]) has been used to determine the adhesive fracture energy during Mode I loading of repair tissue forming between cartilage sample pairs [61]. Cartilage slices, 0.25 mm thick, were harvested from adult bovines, and sample pairs were maintained in complete apposition during 2 weeks of culture with the application of a load to the overlap area (resulting in a normal stress of 0.1 kPa). In the ‘T’ peel test, two flexible adherend sheets (in this case, thin

cartilage blocks) joined together by an adhesive (in this case, the repair tissue) are extended in a ‘T’ configuration to propagate a crack through the adhesive. To apply this to the repaired cartilage sample pair, a small region of the repair tissue between each sample pair was dissected apart to allow the attachment of axles to one end of each cartilage slice. The axles were then subjected to a displacement at a relatively fast rate in order to minimize viscous/frictional energy dissipation. After 2 weeks of incubation in culture medium supplemented with 20% fetal bovine serum, the fracture toughness of the repair tissue, computed as the external work done per unit of crack extension in the plateau region of the measured force, was 16 J/m2 (Table II). These normal and repair fracture properties were within the range of values obtained from Mode I tear tests of other connective tissues, including meniscus, cornea and skin (Table II). Roeddecker et al. [62, 63] studied normal and repaired medial menisci from skeletally immature New Zealand white rabbits. A crack was initiated in the posterior horn of the meniscus and the clamped ends of the specimen were displaced so that the crack propagated circumferentially in the microvascular medial zone towards the anterior horn (Fig. 5C). Continuous load measurements allowed the calculation of tearing energies for di#erent regions along the path of crack propagation, permitting comparison of injured and adjacent non-injured regions. Normalization of the energy values to an assumed meniscal thickness of 1 mm indicated a fracture toughness for normal and repaired rabbit menisci of 1500 J/m2 and 290 J/ m2, respectively (Table II). Comparison to other normal and repair collagenous tissues indicated

36

T. Ahsan and R. L. Sah: Biomechanics of integrative cartilage repair

FIG. 5. Fracture energy testing of cartilaginous tissues. A: Modified single edge notch test of articular cartilage [54]. B: T-peel test of repair tissue forming between two cartilage blocks [61]. C: Top view of tear test of meniscus [63]. D: In-plane shear test of osteochondral junction [52]. E: Trouser tear test of articular cartilage and a representative result ([54], graph reproduced with permission).

that these values were much higher than those from Mode I tear tests of normal human corneal stroma (140 J/m2 [64]), normal and stored rabbit corneal stroma (98 J/m2 [65, 66]), and rat skin graft after 9 days of healing (20 J/m2 [67]). In-plane shear is another mode of loading that may cause failure. Broom et al. [52] assessed the Mode II fracture toughness of normal cartilage at its interface with subchondral bone (i.e., at the osteochondral junction; Fig. 5D). Osteochondral blocks were isolated from the patellae of bovine animals and subjected to impact loading. An instrumented pendulum arm made contact with the sample at the bottom of the swing to cause transverse loading of the cartilage. The impactor was designed to minimize tissue displacement and strain normal to the osteochondral junction by including a constraining face that extended just over the articular surface. From the measured load, energy was computed, and estimates of Mode II fracture toughness for 2–3 year old and 5–7 year old animals averaged 36 and 58 kJ/m2, respectively (Table II). The Mode III fracture toughness of articular cartilage has been estimated using the trouser tear test (Fig. 5E). Chin-Purcell and Lewis [54] used the 0.2 mm thick osteochondral sections, resulting from the MSEN tests described above. The average tearing toughness for canine articular cartilage was calculated from the measured plateau in force,

and varied between 240 and 1200 kJ/m2 (Table II) for samples that fractured relatively sharply and bluntly, respectively, in the MSEN test. By comparison, these values are of the same order of magnitude as the Mode III fracture toughness of pig aorta (1800 J/m2 [68]), and somewhat lower than that of normal rat skin (25 000 J/m2 [69]). Although the rates of energy release during fracture of repair tissue are relatively low compared to that during the fracture of normal tissues, the viscoelastic (biphasic) nature of cartilage complicates the interpretation of these measures of fracture toughness. The measured energy release rates do put an upper bound on the total energy dissipated through fracture and non-fracture mechanisms. Under certain conditions [54], the energy dissipation in a region distant from the crack site will be relatively small. In general, however, regions of the tissue (e.g., at the site of bending) will undergo large deformation, and finite deformation biphasic analysis may be required to evaluate the energy balance in the above test configurations. In addition, the anisotropic, inhomogeneous, fibrous nature leads to preferential failure directions. Cellular and molecular repair mechanisms The biomechanical adhesive properties, described above, provide macroscopic functional

37

Osteoarthritis and Cartilage Vol. 7 No. 1

indices of the integrative cartilage repair process, as well as the integrity of normal cartilage. However, the composition and structure of the repair tissue responsible for these biomechanical properties, as well as the cellular and molecular processes involved, remain to be elucidated. In general, tissue repair and wound healing requires cell migration, adhesion, proliferation, and, ultimately, matrix deposition and remodeling [6, 70, 71]. The initial steps in repair involve the recruitment of an adequate population of cells, often of a specific type, to the defect site. It has been suggested that cell adhesion and migration on cartilage surfaces may be hindered by the anti-adhesive properties of specific matrix molecules [72, 73]. However, other studies indicate that chondrocytes adhere to cut articular cartilage surfaces [74, 75]. The adhesive properties of chondrocytes to cartilage extracellular matrix is unknown, but the adhesive strength of ligament fibroblasts to coated glass surfaces is about 1 kPa [76]. Another possible source of cells in the repair process are the chondrocytes within mature articular cartilage. While these cells may not be able to migrate through the dense extracellular matrix to a defect site, chondrocytes near or at a laceration site requiring integrative repair may be stimulated to undergo limited proliferation [46, 77, 78]. Such proliferation of chondrocytes also can occur within cartilage explants in vitro after enzymatic digestion [79]. It remains to be determined how the density of cells at a certain location, relative to a defect, a#ects the repair process. In the model system of cartilage repair in vitro [42], the repair process did appear to be dependent on cellular activity. Incubation of cartilage explant pairs in the absence of serum partially inhibited the development of adhesive strength, and lyophilization of the cartilage blocks before incubation (in the presence of serum) caused complete inhibition. The development of adhesive strength resulting from the intrinsic biological repair process was not detectably altered by trypsin pre-treatment of the cartilage blocks [80], which has been postulated to promote adhesion of cells exogenous to cartilage [72]. Since the extracellular matrix components are major determinants of the biomechanical function of articular cartilage [e.g., 38, 47], it is likely that the remodeling of specific matrix components is a requirement for successful repair. The large proteoglycan, aggrecan, is known to contribute to the compressive and shear properties of cartilage [38, 47]. It is possible that proteoglycan aggregates can span an interface region and thus contribute to the

integrative repair process. However, since incubation of cartilage explants with IGF-1 alone, but not TGF-â1 alone, stimulated integrative repair in vitro [81], but either IGF-1 and TGF-â1 individually stimulate the synthesis of proteoglycan aggregate in explant culture [82], it seems unlikely that formation of proteoglycan aggregate alone is responsible for the integrative repair process that occurs in bovine cartilage in vitro [42]. The regulation of collagen fibril assembly, disassembly, and reassembly is likely to be critical to integrative cartilage repair. The microscopic analysis by Broom and co-workers [56, 57] indicated that propagation of cracks may occur adjacent to fibers or cause pullout of fibers at the fracture surface, rather than by splitting through fibers. In the 8-week tissues formed by cultured chondrocytes, a linear relation between fracture toughness and collagen content was determined [58]. Since collagen fibrils are of varying size and orientation in the superficial, mid, and deep zones of the articular cartilage [83], the restoration of normal cartilage function may require a complex process of fibril remodeling. The stabilization of heterotypic fibrils of type II, IX, and XI collagen involves crosslinking, mediated by lysyl oxidase [84] and Maillard-glycation [85]. Whether collagen fibril remodeling can be controlled to e#ect an integrative repair process in cartilage, and whether newly synthesized collagen molecules are required for such a process is unknown. The integrative repair of opposing cartilage surfaces or of repair tissue with cartilage is required for cartilage repair. Quantification of the macroscopic biomechanical adhesive properties allows determination of the success of a cartilage repair process. Such measures may help elucidate the cellular and molecular mechanisms involved in repair and the relationships between tissue function, composition, structure and remodeling during repair. Physical, biological, and chemical regulators of the repair process also may be clarified by assessing functional biomechanical indices of integrative repair under di#ering experimental conditions. The determination of basic mechanisms involved in integrative repair and their regulators would help in the design of tissue engineering procedures to stimulate integrative cartilage repair in vivo. Acknowledgments This work was supported by NIH R29 AR44058, NIH P01 AG07996, and NSF BES 9457236. We thank Lisa Lottman for assistance in creating the illustrations, and Dr Jack

38

T. Ahsan and R. L. Sah: Biomechanics of integrative cartilage repair

Lewis and Dr Narendra Simha for constructive critiques of this manuscript.

References 1. Hunter W. On the structure and diseases of articulating cartilage. Phil Trans R Soc London 1743;42:514–21. 2. Brittberg M, Lindahl A, Nilsson A, Ohlsson C, Isaksson O, Peterson L. Treatment of deep cartilage defects in the knee with autologous chondrocyte transplantation. N Engl J Med 1994;331:889– 95. 3. Cartilage repair registry: periodic report. 1997, Genzyme Tissue Repair: Cambridge, MA. 4. Mankin HJ. Chondrocyte transplantation—one answer to an old question. N Engl J Med 1994;331:940–1. 5. Buckwalter JA, Lohmander S. Operative treatment of osteoarthrosis: current practice and future development. J Bone Joint Surg 1994;76A:1405–18. 6. Mankin HJ. The response of articular cartilage to mechanical injury. J Bone Joint Surg 1982;64A:460–6. 7. Bullough PG. The pathology of osteoarthritis. In: Moskowitz RW, et al., Eds. Osteoarthritis: Diagnosis and Medical/Surgical Management, Philadelphia: W.B. Saunders Co. 1992:39–69. 8. Donohue JM, Buss D, Oegema TR, Thompson RC. The e#ects of indirect blunt trauma on adult canine articular cartilage. J Bone Joint Surg 1983;65A:948–57. 9. Thompson RC, Oegema TR, Lewis JL, Wallace L. Osteoarthrotic changes after acute transarticular load: an animal model. J Bone Joint Surg 1991;73A:990–1001. 10. Thompson RC, Vener MJ, Gri$ths HJ, Lewis JL, Oegema TR, Wallace L. Scanning electronmicroscopic and magnetic resonance-imaging studies of injuries to the patellofemoral joint after acute transarticular loading. J Bone Joint Surg 1993;75A:704–13. 11. Sah RL, Amiel D, Coutts RD. Tissue engineering of articular cartilage. Curr Opin Orthop 1995;6:52–60. 12. Fuller JA, Ghadially FN. Ultrastructural observations on surgically produced partial-thickness defects in articular cartilage. Clin Orthop 1972;86:193–205. 13. Ghadially FN, Ailsby RL, Oryschak AF. Scanning electron microscopy of superficial defects in articular cartilage. Ann rheum Dis 1974;33:327–32. 14. Ghadially FN, Thomas I, Oryschak AF, LaLonde J-MA. Long term results of superficial defects in articular cartilage. A scanning electron microscope study. J Pathol 1977;121:213–17. 15. Meachim G. The e#ects of scarification on articular cartilage of rabbit. J Bone Joint Surg 1963;45B:150–61. 16. Meachim G, Roberts C. Repair of the joint surface from subarticular tissue in the rabbit knee. J Anat 1971;109:317–27. 17. Thompson RC. An experimental study of surface injury to articular cartilage and enzyme responses within the joint. Clin Orthop 1975;107:239–48. 18. Messner K, Gillquist J. Cartilage repair. A critical review. Acta Orthop Scand 1996;67:523–9.

19. Shapiro F, Koido S, Glimcher MJ. Cell origin and di#erentiation in the repair of full-thickness defects of articular cartilage. J Bone Joint Surg 1993;75A:532–53. 20. O’Driscoll SW, Keeley FW, Salter RB. Durability of regenerated articular cartilage produced by free autogenous periosteal grafts in major fullthickness defects in joint surfaces under the influence of continuous passive motion. A follow-up report at one year. J Bone Joint Surg 1988;70A:595–606. 21. Ritsila¨ V, Poussa M, Rubak J, Snellman O, O } sterman K. Periosteal and perichondral grafts in reconstruction of the patellar joint surface. Acta Orthop Scand 1980;51:704. 22. Chu CR, Dounchis JS, Yoshioka M, Sah RL, Coutts RD, Amiel D. Osteochondral repair using perichondrial cells: a one year study in rabbits. Clin Orthop Rel Res 1997;340:220–9. 23. Shortkro# S, Barone L, Hsu HP, Wrenn C, Gagne T, Chi T, et al. Healing of chondral and osteochondral defects in a canine model: the role of cultured chondrocytes in regeneration of articular cartilage. Biomaterials 1996;17:147–54. 24. Wakitani S, Goto T, Pineda SJ, Young RG, Mansour JM, Caplan AI, et al. Mesenchymal cell-based repair of large, full-thickness defects of articular cartilage. J Bone Joint Surg 1994;76A:579–92. 25. Cuevas P, Burgos J, Baird A. Basic fibroblast growth factor (FGF) promotes cartilage repair in vivo. Biochem Biophys Res Commun 1988;156:611–18. 26. Helminen HJ, Jurvelin J, Kiviranta I, Paukkonen K, Sa¨a¨ma¨nen A-M, Tammi M. Joint loading e#ects on articular cartilage: a historical review. In: HJ Helminen et al., Eds. Joint Loading: Biology and Health of Articular Structures. Kent: Butterworth 1987:1–46. 27. Guilak F, Sah RL, Setton L. Physical regulation of cartilage metabolism. In: Mow VC, Hayes WC, Eds. Basic Orthopaedic Biomechanics. Philadelphia: Lippincott-Raven 1997:179–207. 28. Eberhardt AW, Lewis JL, Keer LM. Contact of layered elastic spheres as a model of joint contact: e#ect of tangential load and friction. J Biomech Engng 1991;113:107–8. 29. Kelly PA, O’Connor JJ. Transmission of rapidly applied loads through articular cartilage. Part 1: uncracked cartilage. Proc Inst Mech Eng H, J Eng Med 1996;210:27–37. 30. Ateshian GA, Lai WM, Zhu WB, Mow VC. An asymptotic solution for the contact of two biphasic cartilage layers. J Biomechanics 1994;27:1347–60. 31. Mow VC, Bachrach NM, Ateshian GA. The e#ects of a subchondral bone perforation on the load support mechanism within articular cartilage. Wear 1994;175:167–75. 32. Wayne JS, Woo SLY, Kwan MK. Application of the u-p finite element method to the study of articular cartilage. J Biomech Engng 1991;113:397–403. 33. Eberhardt AW, Keer LM, Lewis JL, Vithoontien V. An analytical model of joint contact. J Biomech Engng 1990;112:407–13. 34. Kelly PA, O’Connor JJ. Transmission of rapidly applied loads through articular cartilage. Part 2: cracked cartilage. Proc Inst Mech Eng H, J Eng Med 1996;210:39–49.

39

Osteoarthritis and Cartilage Vol. 7 No. 1

35. Eberhardt AW, Peri S. A vertical surface breaking crack in normal and sliding joint contact. J Biomech Engng 1995;117:153–5. 36. Schinagl RM, Ting MK, Price JH, Sah RL. Video microscopy to quantitate the inhomogeneous equilibrium strain within articular cartilage during confined compression. Ann Biomed Eng 1996;24:500–12. 37. Akizuki S, Mow VC, Muller F, Pita JC, Howell DS, Manicourt DH. Tensile properties of human knee joint cartilage: I. influence of ionic conditions, weight bearing, and fibrillation on the tensile modulus. J Orthop Res 1986;4:379–92. 38. Mow VC, Ratcli#e A. Structure and function of articular cartilage and meniscus. In: Mow VC, Hayes WC, Eds. Basic Orthopaedic Biomechanics, 2nd ed. Philadelphia, PA: Lippincott-Raven Publishers 1997:113–77. 39. Vincent JFV. Biomechanics: Materials. In: Rickwood D, Hames BD, Eds. The Practical Approach Series. New York: IRL Press 1992:247. 40. Adams RD, Wake WC. Structural Adhesive Joints in Engineering. New York: Elsevier Applied Science 1984:309. 41. Stevens KK. Statics and Strength of Materials. Englewood Cli#s, NJ: Prentice-Hall 1987:572. 42. Reindel ES, Ayroso AM, Chen AC, Schinagl RM, Chun DM, Sah RL. Integrative repair of articular cartilage in vitro: adhesive strength of the interface region. J Orthop Res 1995;13:751–60. 43. Ju¨rgensen K, Aeschlimann D, Cavin V, Genge M, Hunziker EB. A new biological glue for cartilagecartilage interfaces: tissue transglutaminase. J Bone Joint Surg 1997;79A:185–93. 44. Jackson RW, Judy MM, Matthews JL, Nosir H. Photochemical tissue welding with 1,8 naphthalimide dyes: in vivo meniscal and cartilage welds (Abstract). Trans Orthop Res Soc 1997;22:650. 45. Standard test method for measuring strength and shear modulus of nonrigid adhesives by the thickadherend tensile-lap specimen. In: American Technical Publishers Ltd, Annual Book of ASTM Standards 1992:312–21. 46. Scully SP, Joyce ME, Heydeman A, Bolander ME. Articular cartilage healing in vitro: modulation by bFGF and TGF-â1 (Abstract). Trans Orthop Res Soc 1991;16:385. 47. Woo SL-Y, Mow VC, Lai WM. Biomechanical properties of articular cartilage. In: Skalak R, Chien S, Eds. Handbook of Bioengineering, New York: McGraw-Hill 1987:4.1–4.44. 48. Woo SL-Y, Akeson WH, Jemmott GF. Measurements of nonhomogeneous directional mechanical properties of articular cartilage in tension. J Biomechanics 1976;9:785–91. 49. Aeschlimann D, Paulsson M. Transglutaminases: protein cross-linking enzymes in tissues and body fluids. Thromb Hemost 1994;71:402–15. 50. Standard test method for shear strength and shear modulus of structural adhesives. In: Annual Book of ASTM Standards, American Technical Publishers Ltd 1992:441–5. 51. Anderson TL. Fracture Mechanics: Fundamentals and Applications. 2nd ed. Boca Raton, FL: CRC Press 1995:688.

52. Broom ND, Oloyede A, Flachsmann R, Hows M. Dynamic fracture characteristics of the osteochondral junction undergoing shear deformation. Med Eng Phys 1996;18:396–404. 53. Kinloch AJ. Adhesion and Adhesives. New York: Chapman and Hall 1987:441. 54. Chin-Purcell MV, Lewis JL. Fracture of articular cartilage. J Biomech Eng 1996;118:545–56. 55. Swanson SR. Introduction to Design and Analysis with Advanced Composite Materials. Upper Saddle River: Prentice-Hall 1997:249. 56. Broom ND, Marra DL. Ultrastructural evidence for fibril-to-fibril associations in articular cartilage and their functional implication. J Anat 1986;146:185–200. 57. Broom ND. Further insights into the structural principles governing the function of articular cartilage. J Anat 1984;139:275–94. 58. Fedewa MM, Leo P, Oegema TR, Lewis JL. E#ect of composition and damage on tensile properties of chondrocyte culture tissue (Abstract). ASME Adv Bioeng 1996;BED33:243–4. 59. Fedewa MM, Lewis JL, Leo P, Oegema TR. Il-1â and retinoic acid degrade both proteoglycan and collagen in chondrocyte culture tissue (Abstract). Trans Orthop Res Soc 1996;21:359. 60. Standard test method for peel resistance of adhesives (T-peel test). In: Annual Book of ASTM Standards, American Technical Publishers Ltd 1992:117–19. 61. Ahsan T, Sah RL. Fracture mechanics characterization of integrative cartilage repair using the T-peel test (Abstract). Trans Orthop Res Soc 1996;21:537. 62. Roeddecker K, Nagelschmidt M, Cutler HM, Jochims J, Muennich U. Tear strength profiles of the meniscus: an animal model for evaluation of injury and therapeutic intervention. Theor Surg 1993;8:136–42. 63. Roeddecker K, Muennich U, Nagelschmidt M. Meniscal healing: A biomechanical study. J Surg Res 1994;56:20–7. 64. Smolek MK, McCarey BE. Interlamellar adhesive strength in human eyebank corneas. Invest Ophthalmol Vis Sci 1990;31:1087–95. 65. Maurice DM, Monroe F. Cohesive strength of corneal lamellae. Exp Eye Res 1990;50:59–63. 66. Barron BC, Smolek MK, McCarey BE. A biomechanical evaluation of rabbit corneas in M-K and Dexsol. Ophthalmic Surg 1992;23:733–7. 67. Dong C, Mead E, Skalak R, Fung YC, Debes JC, Zapata-Sirvent RL, et al. Development of a device for measuring adherence of skin grafts to the wound surface. Ann Biomed Eng 1993;21:51–5. 68. Purslow PP. Positional variations in fracture toughness, sti#ness and strength of descending thoracic pig aorta. J Biomechanics 1983;16:947–53. 69. Purslow PP. Measurement of the fracture toughness of extensible connective tissues. J Mater Sci 1983;18:3591–8. 70. Buckwalter J, Rosenberg L, Coutts R, Hunziker E, Reddi AH, Mow VC. Articular cartilage: injury and repair. In: Woo SLY, Buckwalter JA, Eds. Injury and Repair of the Musculoskeletal Soft Tissues. Rosemont, IL: American Academy of Orthopaedic Surgeons 1988:465–82.

40

T. Ahsan and R. L. Sah: Biomechanics of integrative cartilage repair

71. Cotran RS, Kumar V, Robbins SL. Inflammation and repair. In: Robbins Pathologic Basis of Disease, Ed. Philadelphia, PA: WB Saunders, Co 1989:39– 86. 72. Hunziker EB, Rosenberg LC. Repair of partialthickness defects in articular cartilage: cell recruitment from the synovial membrane. J Bone Joint Surg 1996;78A:721–73. 73. Noyori K, Jasin HE. Inhibition of human fibroblast adhesion by cartilage surface proteoglycans. Arthritis Rheum 1994;37:1656–63. 74. Baragi VM, Renkiewicz RR, Jordan H, Bonadio J, Hartman JW, Roessler BJ. Transplantation of transduced chondrocytes protects articular cartilage from interleukin 1-induced extracellular matrix degradation. J Clin Invest 1995;96:2454–60. 75. Chen AC, Nagrampa JP, Schinagl RM, Lottman LM, Sah RL. Chondrocyte transplantation to articular cartilage explants in vitro. J Orthop Res 1997;15:791–802. 76. Sung K-LP, Kwan MK, Maldonado F, Akeson WH. Adhesion strength of human ligament fibroblasts. J Biomech Eng 1994;116:237–42. 77. Mankin HJ, Boyle CJ. The acute e#ects of lacerative injury on DNA and protein syntheisis in articular cartilage. In: CAL Bassett, Ed. Cartilage Degradation and Repair. Washington, DC: National Research Council–National Academy of Sciences–National Academy of Engineering 1967:185–99. 78. Mankin HJ. Localization of tritiated thymidine in articular cartilage of rabbits. II. repair in immature cartilage. J Bone Joint Surg 1962;44A:682–8. 79. Millroy SJ, Poole AR. Pig articular cartilage in organ culture. e#ect of enzymatic depletion of

80. 81.

82.

83.

84. 85. 86. 87.

the matrix on response of chondrocytes to complement-su$cient antiserum against pig erythrocytes. Ann rheum Dis 1974;33:500–8. Li KW, Ahsan T, Sah RL. E#ect of trypsin treatment on integrative cartilage repair in vitro (Abstract). Trans Orthop Res Soc 1996;21:101 Ahsan T, Li KW, Lau ST, Sah RL. Modulation of integrative cartilage repair in vitro by IGF-1 and TGF-â1 (Abstract). Trans Orthop Res Soc 1995;20:173. Morales TI, Hascall VC. Transforming growth factor-â1 stimulates synthesis of proteoglycan aggregates in calf articular organ cultures. Arch Biochem Biophys 1991;286:99–106. Birk DE, Linsenmayer TF. Collagen fibril assembly, deposition, and organization into tissue-specific matrices. In: Yurchenco PD, Birk DE, Mecham RP, Eds. Extracellular Matrix Assembly and Structure. San Diego, CA: Academic Press 1994: 91–128. Eyre DR, Paz MA, Gallop PM. Cross-linking in collagen and elastin. Ann Rev Biochem 1984;53:717–48. Reiser K, McCormick RJ, Rucker RB. Enzymatic and nonenzymatic cross-linking of collagen and elastin. FASEB J 1992;6:2439–49. Smolek MK. Interlamellar cohesive strength in the vertical meridian of human eyebank corneas. Invest Ophthalmol Vis Sci 1993;34:2962–9. Koutroupi KS, Barbenel JC. Mechanical and failure behavior of the stratum corneum. J Biomechanics 1990;23:281–7.