Tendons and ligaments are both dense, regularly arranged connective tissues, but the collagen fibers in a tendon are more parallel to the longitudinal axis than is the case in a ligament. The collagen fibers, composed of thinner fibrils, extend the entire length of the tendon.
The surface of the tendon is enveloped in a white, glistening, synovial-like membrane, called the epitenon , which is continuous on its inner surface with the endotenon , a thin layer of connective tissue that binds collagen fibers and contains lymphatics, blood vessels, and nerves. In some tendons, the epitenon is surrounded by a loose areolar tissue called the paratenon that functions as an elastic sheath through which the tendon can slide. In some tendons, the paratenon is replaced by a true synovial sheath or bursa consisting of two layers lined by synovial cells, called the tenosynovium , within which the mesotendon carries important blood vessels to the tendon. In the absence of a synovial lining, the paratenon often is called a tenovagina . Together the epitenon and the paratenon compose the peritenon ( Fig. 1-1 ). The blood supply to tendon has several sources, including the perimysium, periosteal attachments, and surrounding tissues. Blood supplied through the surrounding tissues reaches the tendon through the paratenon, mesotenon, or vincula. Vascular tendons are surrounded by a paratenon and receive vessels along their borders; these vessels then coalesce within the tendon. The relatively avascular tendons are contained within tendinous sheaths, and the mesotenons within these sheaths function as vascularized conduits called vincula . The muscle-tendon and tendon-bone junctions, along with the mesotenon, are the three types of vascular supply to the tendon inside the sheath. Other sources of nutrition include diffusional pathways from the synovial fluid, which provide an important supply of nutrients for the flexor tendons of the hand, for example. The nervous supply to a tendon is sensory in nature. The proprioceptive information supplied to the central nervous system by these nerves usually is picked up through mechanoreceptors located near the musculotendinous junction.
The tendon-bone interface marks the site where collagen fibers enter bone as Sharpey fibers, and the endotenon becomes continuous with the periosteum. The insertion of tendon into bone generally is classified into two types. The simpler type, termed direct insertion , occurs when the tendon fibrils pass directly into bone through zones of fibrocartilage with little interdigitation into the surrounding periosteum. The tendon inserts into a zone of fibrocartilage, then into a layer of mineralized fibrocartilage, and finally into bone. Dissipation of force is achieved effectively through this gradual transition from tendon to fibrocartilage to bone. The second type of insertion is more complex; the superficial fibrils insert into the periosteum, whereas the deeper fibrils fan out into bone directly. When tendons insert at an angle into the bone, a larger area of fibrocartilage can be found on one side of the insertion, which is thought to be an adaptation to the compressive forces experienced by the tendon on that side.
Tendons consist of relatively few cells and an abundant extracellular matrix. The cellular component of tendon is the tenocyte; the main constituent of the matrix is collagen, along with small amounts of elastin, ground substance, and water.
Because of the arrangement of collagen fibers parallel to the tendon long axis, tendons have one of the highest tensile strengths of all soft tissues. All types of collagen have in common a triple helical domain, which is combined differently with globular and nonhelical structural elements. The triple helix conformation of collagen is stabilized mainly by hydrogen bonds between glycine residues and between hydroxyl groups of hydroxyproline. This helical conformation is reinforced by hydroxyproline-forming and proline-forming hydrogen bonds to the other two chains. The physical properties of collagen and its resistance to enzymatic and chemical breakdown rely on covalent cross-links within and between the molecules.
Elastin is a protein that allows connective tissues to undergo large changes in geometry while expending little energy in the process. Tendons of the extremities possess small amounts of this structural protein, whereas elastic ligaments, such as the ligamentum flavum and ligamentum nuchae, have greater proportions of elastin. In most tendons, elastin is found primarily at the fascicle surface comprising less than 1% of the tendon by dry weight, and it is responsible for the crimp pattern of the tendon when viewed by a light microscope. Elastin is similar to collagen in that it has lysine-derived cross-links. The amino acids desmosine and isodesmosine are unique to elastin. Their formation depends on the presence of copper. The elastic potential of elastin is primarily attributed to the cross-linking of lysine residues through desmosine, isodesmosine, and lysin norleucine.
Approximately 1% of the total dry weight of tendon is composed of ground substance, which consists of proteoglycans, glycosaminoglycans, structural glycoproteins, plasma proteins, and a variety of small molecules. Proteoglycans and glycosaminoglycans are thought to be important for stabilizing the collagenous skeleton of connective tissue. Regions of tendon that experience primarily tensile forces have a lower proteoglycan content and higher rates of collagen synthesis than do areas that experience frictional and compressive forces in addition to tensile forces.
When the mechanical forces on the tendon exceed the maximum strain or stress that the tissue can accept, injury occurs. Injury to tendons can result from acute trauma (e.g., laceration) or repetitive loading (e.g., overuse injury).
Controversy exists in the literature about a universal classification of overuse tendon injuries and the pathologic entities responsible for them. A classification of Achilles tendon disorders provides a guide to the structural manifestations of overuse injury as follows: (1) peritendinitis, or inflammation of the peritenon; (2) tendinosis with peritendinitis; (3) tendinosis without peritendinitis; (4) partial rupture; and (5) total rupture. Other classifiers have added a sixth category, tendinitis, in which the primary site of injury is the tendon, with an associated reactive peritendinitis. The classification is not universal because some tendons lack a paratenon and instead have synovial sheaths; furthermore, it is unclear if certain histopathologic conditions are actually separate entities. For instance, human biopsy studies have been unable to show histologic evidence of inflammation within the tendon substance. Because of uncertainty regarding the histologic features of these conditions, several authors have suggested use of the term tendinopathy rather than tendinitis .
Spontaneous tendon rupture sometimes occurs during sporting activities even though laboratory studies have shown that under normal circumstances healthy tendon is not the weak link of the musculotendinous unit. Healthy tendon is in fact stronger than its muscle or muscle-tendon junction. Biopsies have shown that preexisting degenerative changes are present in most tendons that spontaneously rupture. Histologic tendon degeneration can occur without clinical symptoms. Thus it appears that preexisting but generally asymptomatic pathology precedes spontaneous tendon rupture during sports-related activity.
Studies have shown that in cases of chronic tendon pain the pathologic lesion is typical of a degenerative process rather than an inflammatory one and that this degeneration occurs in areas of diminished blood flow. Several authors have documented the existence of areas of marked degeneration without acute or chronic inflammation in most of these cases. These changes are separate and distinct from the site of rupture. A review of patients with chronic tendinitis syndrome revealed similar findings of tendon degeneration. Nirschl described the pathology of chronic tendinitis as angiofibroblastic hyperplasia. A characteristic pattern of fibroblasts and vascular, atypical, granulation-like tissue can be seen microscopically. Cells characteristic of acute inflammation are virtually absent. These observations suggest that factors other than mechanical overuse play an important role in the pathogenesis of these tendon lesions.
In several studies a correlation between the incidence of chronic tendon problems and increased age has been identified. In vitro studies have shown decreased proliferative and metabolic responses of aging tendon tissue. Other causative factors include the lack of blood flow in certain areas (supraspinatus and Achilles tendon, for example) that may predispose a tendon to rupture or may result in chronic tendinopathy. Biopsy specimens of young patients with chronic tendinitis have revealed a change in the morphology of tenocytes adjacent to areas of collagen degeneration. Tendon compression may also have a role in insertional tendinopathy. For example, compression of the tendon at the origin on the inferior pole of the patella may play a role in patellar tendinopathy, or jumper’s knee.
Two different theories of primary tendon healing exist. One theory suggests that healing depends on the surrounding tissues and that the tendon itself plays no significant role. This theory holds that the tendon is an inert, almost avascular structure the cells of which are incapable of contributing to the healing process. Potenza showed that the tendon is invaded by fibrovascular tissue at the location of suture placement. At 28 days, the collagen produced by these fibroblasts is immature, but by 128 days, it is indistinguishable from that of normal tendon. In contrast, several studies have suggested that the inflammatory response is not essential to the healing process and that tendons possess an intrinsic capacity for repair. Lindsay and Thomson were the first to show that an experimental tendon suture zone can be isolated from the perisheath tissues and that healing progressed at the same rate as when the perisheath tissues were intact. Later, in isolated segments of profundus tendon in rabbits, these researchers found anabolic and catabolic enzymes, which showed that an active metabolic process existed in the experimentally free tendons. In addition, sutured free tendon grafts of rabbit flexor tendons healed without adhesions within a vascular synovial environment of the suprapatellar bursa. Consequently, it is now accepted that tendons may possess intrinsic and extrinsic capabilities for healing and that the contribution of each of these two mechanisms probably depends on the location, extent, and mechanism of injury and the rehabilitation program used after the injury.
As in other areas in the body, tendon healing proceeds in three phases: (1) an inflammatory stage, (2) a reparative or collagen-producing stage, and (3) a remodeling phase.
Tendon healing begins with the formation of a blood clot and an inflammatory reaction that includes an outpouring of fibrin and inflammatory cells. The degree of inflammation is related to the size of the wound and the amount and type of trauma that has occurred. The presence of nitric oxide also appears to limit the duration and intensity of the inflammatory response after a tendon injury. A clot forms between the two tendon ends and is invaded by cells resembling fibroblasts and migratory capillary buds. The fibroblasts are believed to arise from the endotenon and epitenon. Fibroblasts residing in the endotenon differ from those in the epitendinous tissues. The epitendinous fibroblasts resemble synovial cells and are particularly active in response to injury. Mesenchymal cells, most likely from circulation, which are capable of differentiating into fibroblasts, also appear in the area during the healing process. In the subsequent days and weeks, the number of circulation-derived mesenchymal cells decreases and that of locally derived mesenchymal cells from the tendon tissue itself increases. The inflammatory phase is evident until the eighth to tenth day after injury.
Collagen synthesis begins within the first week and reaches its maximal level after about 4 weeks. At 3 months, collagen synthesis continues at a rate three to four times that of normal tissue. Type I collagen is synthesized and extruded into the extracellular space as procollagen, which is converted to type I collagen by the enzyme procollagenase. The expression of procollagen messenger ribonucleic acid, and thus the production of procollagen, depends on the presence of transforming growth factor (TGF)- β 1 ; other growth factors are also expressed throughout the acute inflammatory phase. Initially, the collagen fibrils are oriented perpendicular to the long axis of the tendon, but by 2 months these fibrils usually are oriented parallel to the axis of tensile loading.
Restoration of the gliding function of the tendon depends on the dissolution and reformation of the collagen fibers during the scar remodeling phase. This phase starts at about the fifteenth day, and by 28 days, most of the fibroblasts and collagen between the tendon stumps are oriented longitudinally. Collagenase is present in the wound on the second day after injury. Between 4 and 6 weeks, collagen synthesis and collagen degradation reach equilibrium. Collagen maturation and remodeling begin in the third week and can continue for 1 year after the injury occurs. The strength of the tendon repair results from the organization of collagen fibrils at the bone site; these fibrils cross-link with each other and with those of the tendon on each side of the wound.
The strength of an injured tendon that has been sutured properly increases rapidly during the fibroplastic phase, when granulation tissue is produced to repair the defect. Quantitative changes in acid mucopolysaccharides (hydroxyproline and hexosamine) accompany collagen production, and the ratio of wound collagen to mucopolysaccharide content is a direct measure of increasing tensile strength. The strength of the healing tendon increases as the collagen becomes stabilized by cross-links and the fibrils assemble into fibers. During the maturation phase, the mechanical strength of the healing tendon increases as a result of remodeling and reorganization of the fiber architecture.
Factors Affecting Healing
Application of Load
Active mobilization in the immediate postoperative period (<3 weeks after surgery) may have a deleterious effect on tendon healing by increasing tension across the suture line, leading to gap formation and tendon ischemia, tenomalacia, and possible tendon rupture. A study of rats with collagenase-induced Achilles tendon injuries found that exercising immediately after injury increased and prolonged the presence of inflammatory cells (neutrophils and macrophages) and decreased the stiffness and ultimate force. However, after 3 weeks of immobilization, once the fibrous union has strengthened, active mobilization can stimulate healing, leading to a threefold increase in tendon tensile strength with mechanical stress promoting alignment of the collagen fibrils. Delayed loading can also enhance the mechanical and biologic characteristics of tendon-to-bone healing compared with immediate loading or prolonged postoperative immobilization. The optimal amount of tension necessary to promote an acceptable clinical response is currently unknown, but it is clear that remodeling of collagen scar tissue into mature tendon tissue depends on the presence of tensile forces. The concept of immediate passive mobilization was introduced by Kleinert and coworkers, who showed that during limited active extension reciprocal relaxation of the flexor tendons occurs, allowing passive extension of the repaired tendon. This controlled passive motion was found to be effective experimentally and clinically in decreasing the tethering effect of adhesions and in improving the rates of tendon repair, gliding function, and strength of the tendon. Indeed, from a biomechanical point of view, the optimal procedure appears to be to use a strong suture repair to reduce gap formation and scar tissue in the early phase of healing, after which tensile stress placed on the tendon through controlled passive mobilization can be used to promote earlier reorganization and remodeling of the collagen, leading to achievement of higher tensile strength.
Effect of Corticosteroids and Nonsteroidal Antiinflammatory Drugs on the Injured Tendon
Large doses of corticosteroids inhibit wound healing, but study findings on the effects of small and moderate doses are conflicting. Experimental data on the healing of injured tendons are lacking. Studies of other fibroblastic structures have shown that corticosteroids suppress the formation of adhesions but may lower the tensile strength of sutured tendons, leading to subsequent spontaneous rupture. For example, rats injected with cortisone have increased stiffness in 10-day wounds but decreased energy absorption before tendon failure. Investigators using a rat model found that the administration of either indomethacin, a nonselective antiinflammatory drug, or celecoxib, a cyclooxygenase-2–specific antiinflammatory drug, inhibited tendon-to-bone healing. The authors concluded that interfering with cyclooxygenase-2 activity in the early inflammatory cascade has an adverse effect on tendon-to-bone healing. A study of rotator cuff repair in sheep showed that the stiffness of the repair construct during this stage can be improved by augmenting the repair with a collagen patch (swine small intestine submucosa [SIS]), although the patch had no effect on load to failure. The long-term effect of xenograft collagen patches on the rate and quality of healing in rotator cuff repair remains uncertain.
Methods for Augmentation of Tendon Healing
A large body of research has demonstrated the potential for growth factors to improve tendon tissue healing by stimulation of cell proliferation, chemotaxis, matrix synthesis, and cell differentiation (summarized in Table 1-1 ). However, the challenge at this time is to identify the optimal carrier vehicles to deliver cytokines to the desired site for a relevant period. Furthermore, the complexities of tendon healing will likely require a cascade of various cytokines, delivered at different times, suggesting a distinct limitation of single-factor therapy. Platelet-rich plasma (PRP) can be used to deliver a “cocktail” of cytokines; however, the results of PRP for augmentation of tendon healing have been variable, likely because of the tremendous variability in different PRP formulations. In addition to multifunctional cytokines such as TGF-β and platelet-derived growth factor, recent work has focused on recapitulating the cellular and molecular signals that are expressed during embryonic tendon development. Examples include signaling molecules such as scleraxis and TGF-β3.
Cell-based approaches also appear promising for tendon tissue engineering and improvement of tendon healing. Pluripotent stem cells are currently being evaluated in tendon repair. PRP and cytokines provide the “signals”; however, an appropriate cell population also likely is needed to regenerate the microstructure of normal tendon. The challenge at this time is to identify the optimal cell source and to further elucidate how to drive differentiation of these cells into mature tenocytes.
Recent research has also investigated scaffold materials to augment tendon repair. Porcine-derived SIS has been used as a scaffold to augment tendon repair. However, an excessive inflammatory/immunologic response to the SIS material has been reported, which is believed to be due to residual porcine DNA in the implant. Clinical results using SIS for tendon repair have been poor. Other scaffold materials that contain collagen are being evaluated for their role in augmentation of tendon healing.
Ligaments grossly appear as firm, white fibrous bands, sheets, or thickened strips of joint capsule securely anchored to bone. They consist of a proximal bone insertion, the substance of the ligament or the capsule, and a distal bone insertion. Because most insertions are no more than 1 mm thick, they contribute only a small amount to the volume and the length of the ligament. Bundles of collagen fibrils form the bulk of the ligament substance. Some ligaments consist of more than one band of collagen fibril bundles. For example, the anterior cruciate ligament has a continuum of fiber lengths; different fibers become taut throughout the range of motion. The alignment of collagen fiber bundles within the ligament substance generally follows the lines of tension applied to the ligament during normal activities. In addition, light microscopic examination has shown that the collagen bundles have a wave or crimp pattern. The crimp pattern of matrix organization may allow slight elongation of the ligament without incurring damage to the tissue. In some regions, the ligament cells align themselves in rows between collagen fiber bundles, but in other regions, the cells lack apparent orientation relative to the alignment of the matrix collagen fibers. Scattered blood vessels penetrate the ligament substance, forming small-diameter, longitudinal vascular channels that lie parallel to the collagen bundles. Nerve fibers lie next to some vessels, and nerve endings with the structure of mechanoreceptors have been found in some ligaments.
Ligament insertions vary in size, strength, angle of the ligament collagen fiber bundles relative to the bone, and proportion of ligament collagen fibers that penetrate directly into bone. Based on the angle between the collagen fibrils and the bone and the proportion of the collagen fibers that penetrate directly into bone, investigators group ligament insertions into two types: direct and indirect.
Direct ligament insertions consist of sharply defined regions where the ligament appears to pass directly into the cortex of the bone. The thin layer of superficial ligament collagen fibers of direct insertions joins the fibrous layer of the periosteum. Most of the ligament insertion consists of deeper fibers that directly penetrate the cortex, often at a right angle to the bone surface. The deeper collagen fibers pass through four zones with increasing stiffness: ligament substance, fibrocartilage, mineralized fibrocartilage, and bone. Indirect or oblique ligament insertions into bone, such as the tibial insertion of the medial collateral ligament of the knee or the femoral insertion of the lateral collateral ligament, are less common than are direct insertions. They usually cover more bone surface area than direct insertions, and their boundaries cannot be easily defined because the ligament passes obliquely along the bone surface rather than directly into the cortex.
Fibroblasts are the dominant cells of ligaments, although endothelial cells of small vessels and nerve cell processes are also present. Fibroblasts form and maintain the extracellular matrix. They vary in shape, activity, and density among ligaments, among regions of the same ligament, and with the age of the tissue. Many fibroblasts are spindle shaped and extend between the collagen fibrils. Tissue fluid contributes 60% or more of the wet weight of most ligaments. Because many ligament cells lie at some distance from vessels, these cells must depend on the diffusion of nutrients and metabolites through the tissue fluid. The interaction of the tissue fluid and the matrix macromolecules influences the mechanical properties of the tissue. Fibrillar collagen has the form of cylindrical cross-banded fibrils when examined with electron microscopy. These fibrils give ligaments their form and their tensile strength and constitute 70% to 80% of the dry weight of the ligament.
Type I collagen, which is the major component of the molecular framework, composes more than 90% of the collagen content of ligaments. Type III collagen constitutes about 10% of the collagen, and small amounts of other collagen types may be present as well. Ligaments have more type III collagen than do tendons.
Most ligaments have little elastin (usually less than 5%), but a few, such as the nuchal ligament and the ligamentum flavum, have high concentrations (up to 75%). Elastin forms protein fibrils or sheets, but elastin fibrils lack the cross-banding pattern of fibrillar collagen and differ in amino acid composition, including two amino acids not found in collagen (desmosine and isodesmosine). Also unlike collagen, elastin amino acid chains form random coils when the molecules are unloaded. This conformation of the amino acid chains makes it possible for elastin to undergo some deformation without rupturing or tearing and then, when the load is removed, to return to its original size and shape.
Proteoglycans form only a small portion (less than 1% dry weight) of the macromolecular framework of the ligament but may have important roles in organizing the extracellular matrix and interacting with the tissue fluid. Most ligaments have a higher concentration of glycosaminoglycans than do tendons. Like tendon, meniscus, and articular cartilage, ligaments contain two known classes of proteoglycans. Larger, articular cartilage-type proteoglycans contain long negatively charged chains of chondroitin and keratan sulfate. Smaller proteoglycans contain dermatan sulfate. Because of their long chains of negative charges, the large articular cartilage-type proteoglycans tend to expand to their maximal domain in solution until restrained by the collagen fibril network. As a result, they maintain water within the tissue and exert a swelling pressure, thereby contributing to the mechanical properties of the tissue and filling the regions between the collagen fibrils. The small leucine-rich proteoglycans usually lie directly on the surface of collagen fibrils and appear to affect formation, organization, and stability of the extracellular matrix, including collagen fibril formation and diameter. They may also control the activity of growth factors by direct association.
Although noncollagenous proteins contribute only a small percentage of the dry weight of dense fibrous tissues, they appear to help organize and maintain the macromolecular framework of the collagen matrix, aid in the adherence of cells to the framework, and possibly influence cell function. One noncollagenous protein, fibronectin, has been identified in the extracellular matrix of ligaments and may be associated with several matrix component molecules and with blood vessels. Other noncollagenous proteins undoubtedly exist within the ligament matrix, but their identity and their functions have not yet been defined. Many of the noncollagenous proteins also contain a few monosaccharides and oligosaccharides.
Ligament strains and tears disrupt the matrix, damage blood vessels, and injure or kill cells. Damage to cells, matrices, and blood vessels and the resulting hemorrhage start a response that includes inflammation, repair, and remodeling. These events form a continuous sequence of cell, matrix, and vascular changes that begins with the release of inflammatory mediators and ends when remodeling ceases.
Acute inflammation lasts 48 to 72 hours after most ligament injuries and then gradually resolves as repair progresses. Some of the events that occur during inflammation, including the release of cytokines or growth factors, may help stimulate tissue repair. These mediators promote vascular dilation and increase vascular permeability, leading to exudation of fluid from vessels in the injured region, which causes tissue edema. Ligament tissue immediately surrounding the injury becomes swollen and increasingly friable. Uninjured ligament tissue at a distance from the injury also swells as a result of exudation of fluid from the dilated vessels. Blood escaping from the damaged vessels forms a hematoma that temporarily fills the injured site. Fibrin accumulates within the hematoma, and platelets bind to fibrillar collagen, thereby achieving hemostasis and forming a clot consisting of fibrin, platelets, red cells, and cell and matrix debris. The clot provides a framework for vascular and fibroblast cell invasion. As they participate in clot formation, platelets release vasoactive mediators and the cytokines or growth factors (TGF-β and platelet-derived growth factor).
Within hours of the injury, polymorphonuclear leukocytes appear in the damaged tissue and the clot. Shortly thereafter, monocytes arrive and increase in number until they become the predominant cell type. Enzymes released from the inflammatory cells help digest necrotic tissue, and monocytes phagocytose small particles of necrotic tissue and cell debris. Endothelial cells near the injury site begin to proliferate, creating new capillaries that grow toward the region of tissue damage. Release of chemotactic factors and cytokines from endothelial cells, monocytes, and other inflammatory cells helps to stimulate migration and proliferation of the fibroblasts that begin the repair process.
Ligament repair involves the replacement of necrotic or damaged tissue by cell proliferation and synthesis of new matrix. Repair depends on the fibroblasts that migrate into the injured tissue and clot. Within 2 to 3 days of the injury, fibroblasts within the wound begin to proliferate rapidly and synthesize new matrix. They replace the clot and the necrotic tissue with a soft, loose fibrous matrix containing high concentrations of water, glycosaminoglycans, and type III collagen. Inflammatory cells and fibroblasts fill this initial repair tissue. Within 3 to 4 days, vascular buds from the surrounding tissue grow into the repair tissue and then canalize to allow blood flow to the injured tissue and across small tissue defects. This vascular granulation tissue fills the tissue defect and extends for a short distance into the surrounding tissue but has little tensile strength.
During the next several weeks, as repair progresses, the composition of the granulation tissue changes. Water, glycosaminoglycan, and type III collagen concentrations decline, the inflammatory cells disappear, and the concentration of type I collagen increases. Newly synthesized collagen fibrils increase in size and begin to form tightly packed bundles, and the density of fibroblasts decreases. Matrix organization increases as the fibrils begin to align along the lines of stress, the number of blood vessels decreases, and small amounts of elastin may appear within the site of injury. The tensile strength of the repair tissue increases as the collagen concentration increases.
Repair of many ligament injuries results in an excessive volume of highly cellular tissue with limited mechanical properties and a poorly organized matrix. Remodeling reshapes and strengthens this tissue by removing, reorganizing, and replacing cells and matrix. In most ligament injuries, evidence of remodeling appears within several weeks of injury as fibroblasts and macrophages decrease, fibroblast synthetic activity decreases, and fibroblasts and collagen fibrils assume a more organized appearance. As these changes occur in the repair tissue, collagen fibrils grow in diameter, the concentration of collagen and the ratio of type I to type III collagen increase, and the water and proteoglycan concentrations decline.
In the months after the injury occurs, the matrix continues to align, presumably in response to loads applied to the repair tissue. The most apparent signs of remodeling disappear within 4 to 6 months of injury. The cell density and the number of small blood vessels decline to near-normal levels, and the collagen concentration increases nearly to the level of that in normal tissue. Removal, replacement, and reorganization of repair tissue, however, continue to some extent for years.
Factors Affecting Healing
Size, Location of Defect, and Ligament Type
Among the most important variables that affect healing of the ligament are the type of ligament, the size of the tissue defect, and the amount of load applied to the ligament repair tissue. For example, injuries to capsular and extracapsular ligaments stimulate production of repair tissue that will fill most defects, but injuries to intracapsular ligaments, such as the anterior cruciate ligament, often fail to produce a successful repair response. Treatments that achieve or maintain apposition of torn ligament tissue and that stabilize the injury site decrease the volume of repair tissue necessary to heal the injury, which can benefit the healing process. Such treatments may also minimize scarring and help provide near-normal tissue length. For these reasons, avoidance of wide separation of ruptured ligament ends and selection of treatments that maintain some stability at the injured site during the initial stages of repair are generally desirable. Early controlled loading of ligament repair tissue can promote healing, but excessive loading will disrupt repair tissue and delay or prevent healing.
Methods for Augmentation of Ligament Healing
Some of the same approaches that are being investigated for tendon repair, such as cytokines, gene transfer, stem cells, and scaffolds, are also being tested for augmentation of ligament healing. Cytokines may improve ligament healing by stimulation of cell proliferation, chemotaxis, matrix synthesis, and cell differentiation. However, the challenge at this time is to identify the optimal carrier vehicles to deliver cytokines to the desired site for a relevant period. PRP is a method to deliver autologous cytokines; however, the results of PRP for augmentation of ligament healing have been variable, likely because of the tremendous variability in different PRP formulations. An appropriate cell population is likely also needed to regenerate the microstructure of normal ligament. Ultimately, cell-based approaches combined with cytokines hold promise for improving ligament healing.
Because ligament reconstruction often requires use of a tendon graft, current research is focusing on tissue engineering techniques to create functional tissue replacements. Nanomaterials are promising for ligament tissue engineering, because the microstructure of the material mimics native extracellular matrix. Multiphasic scaffolds are being used to create bone-ligament composites. In addition to various scaffold materials and cell types, it has become clear that mechanical stimulation of the neotissue is also critical to optimize the structure and composition of the tissue. The specific scaffold can be modified in vitro by seeding marrow stromal cells on the scaffold and applying cyclic stretching to increase the alignment of cells, as well as to improve the production and orientation of collagen. When applied in vivo, such a tissue-engineered scaffold could serve to accelerate the healing process, ultimately helping to make a better neoligament or tendon.
Human menisci are semilunar in shape and consist of a sparse distribution of cells surrounded by an abundant extracellular matrix. Within the meniscus lies an anisotropic, inhomogeneous, and highly ordered arrangement of collagen fibrils. The meniscal surface consists of a randomly woven mesh of fine collagen type II fibrils that lie parallel to the surface. Below this surface layer, large, circumferentially arranged collagen fiber bundles (mostly type I) spread through the body of the tissue ( Fig. 1-2 ). These circumferential collagen bundles give menisci great tensile stiffness and strength parallel to their orientation. The collagen bundles insert into the anterior and the posterior meniscal attachment sites on the tibial plateau, providing for stiff and strong attachment sites. Figure 1-2, A , illustrates these large fiber bundles and the thin superficial surface layer. Figure 1-2, B , is a photograph of a bovine medial meniscus with the surface layer removed, showing the large collagen bundles of the deep zone.
Radial sections of meniscus ( Fig. 1-3 ) show radially oriented bundles of collagen fibrils, or “radial tie fibers,” among the circumferential collagen fibril bundles, weaving from the periphery of the meniscus to the inner region. Presumably, the tie fibers help increase the stiffness and the strength of the tissue in a radial direction, thereby resisting longitudinal splitting of the collagen framework. In cross section, these radial tie fibers appear to be more abundant in the middle and the posterior sections than in the anterior sections of the meniscus.
Unlike articular cartilage, the peripheral 25% to 30% of the lateral meniscus and the peripheral 30% of the medial meniscus have a blood supply, and the peripheral regions of the meniscus, especially the meniscal horns, have a nerve supply. Branches from the geniculate arteries form a capillary plexus along the peripheral borders of the menisci. Small radial branches project from these circumferential parameniscal vessels into the meniscal substance.
The mechanical functions of the menisci depend on a highly organized extracellular matrix consisting of fluid and a macromolecular framework formed of collagen (types I, II, III, V, and VI), proteoglycans, elastin, and noncollagenous proteins, along with the cells that maintain this matrix.
Based on morphologic characteristics, two major types of meniscal cells exist. Near the surface, the cells have flattened ellipsoid or fusiform shapes; in the deep zone, the cells are spherical or polygonal. The superficial and the deep meniscal cells appear to have different metabolic functions and perhaps different responses to loading. Like most other mesenchymal cells, these cells lack cell-to-cell contacts. Because most of the cells lie at a distance from blood vessels, they rely on diffusion through the matrix for transport of nutrients and metabolites. The membranes of meniscal cells attach to matrix macromolecules through adhesion proteins (e.g., fibronectin, thrombospondin, and type VI collagen ). The matrix, particularly the pericellular region, protects the cells from damage due to physiologic loading of the tissue. Deformation of the macromolecular framework of the matrix causes fluid flow through the matrix and influences meniscal cell function. Because meniscal tissue is more fibrous than hyaline cartilage, some authors have proposed that meniscal cells be called fibrochondrocytes. Water contributes 65% to 75% of the total weight of the meniscus. Some portion of the water may reside within the intrafibrillar space of the collagen fibers. Most of the water is retained within the tissue in the solvent domains of the proteoglycans by means of both their strong hydrophilic tendencies and the Donnan osmotic pressure exerted by the counter ions associated with the negative charge groups on the proteoglycans. Because the pore size of the tissue is extremely small (<60 nm), very large hydraulic pressures are required to overcome the drag of frictional resistance when forcing fluid flow through the tissue. These interactions between water and the macromolecular framework of the matrix significantly influence the viscoelastic properties of the tissue.
Some meniscal regions have a proteoglycan concentration of up to 3% of their dry weight. Like proteoglycans from other dense fibrous tissues, meniscus proteoglycans can be divided into two general types. The large, aggregating proteoglycans expand to fill large volumes of matrix and contribute to tissue hydration and the mechanical properties of the tissue. The smaller, nonaggregating proteoglycans usually have a close relationship with fibrillar collagen. The large aggregating proteoglycans from meniscus have the same structure as the large aggregating proteoglycans from articular cartilage. The concentration of large aggregating proteoglycans suggests that they probably contribute less to the properties of meniscus than to the properties of articular cartilage. As with the quantitatively minor collagens, the smaller nonaggregating meniscal proteoglycans may help organize and stabilize the matrix, but at present, their function remains unknown.
Noncollagenous proteins also form part of the macromolecular framework of the meniscus and may contribute as much as 10% of the dry weight of the tissue in some regions. Two specific noncollagenous proteins, link protein and fibronectin, have been identified in the meniscus. Link protein is required for the formation of the stable proteoglycan aggregates that are capable of forming strong networks. Fibronectin serves as an attachment protein for cells in the extracellular matrix. Other noncollagenous proteins such as thrombospondin may serve as adhesive proteins in the tissue, thus contributing to the structure and the mechanical strength of the matrix. The exact details of their composition and function in the meniscus remain largely unknown. Finally, elastin contributes less than 1% of the dry weight of the meniscus. The contribution of elastin to the mechanical properties of meniscal tissue is uncertain. The sparsely distributed elastic fibers are unlikely to play a significant role in the organization of the matrix or in determining the mechanical properties of the tissue.
Traumatic meniscal tears occur most frequently in young, active people. Tension, compression, or shear forces that exceed the strength of the meniscal matrix in any direction can lead to tissue failure. Acute traumatic injuries of normal meniscal substance usually produce longitudinal or transverse tears, although the morphology of tears can be quite complex. The configuration of tears due to overloading of normal meniscal tissue depends strongly on the direction and the rate of stretch. Unlike acute traumatic tears through apparently normal meniscal tissue, degenerative meniscal tears occur in association with age-related changes in the tissue. These degenerative tears are most common in persons older than 40 years. Often, these persons do not recall a specific injury, or they recall only a minor load applied to the knee. Degenerative tears often have complex shapes or may appear as horizontal clefts or flaps, as though they were produced by shear failure. Multiple degenerative tears often occur within the same meniscus. These features of degenerative meniscal tears suggest that they result more from age-related changes in the collagen-proteoglycan solid matrix than from specific acute trauma.
The response of meniscal tissue to tears depends on whether the tear occurs through a vascular or an avascular portion of the meniscus. The vascular regions respond to injury as other vascularized dense fibrous tissues do. The tissue damage initiates a sequence of cellular and vascular events recognized as inflammation, repair, and remodeling that can result in healing and restoration of tissue structure and function. Tears through the vascular regions of the meniscus can heal, but tears through the avascular regions do not undergo a repair process, resulting in tissue deficiency.
Factors Affecting Healing
Repair in Vascular Regions of the Meniscus
A tear in the vascular region of the meniscus damages blood vessels, causing hemorrhage and fibrin clot formation. Injury to meniscal cells and the process of clot formation activate inflammation. As in other vascularized dense fibrous tissues (e.g., tendon and ligament), mesenchymal cells and vascular buds invade the fibrin clot. Platelets within the clot and inflammatory cells release mediators that stimulate cell migration, proliferation, and differentiation. If the injury site is sufficiently stable to allow repair, granulation tissue replaces the fibrin clot, and vessels soon cross the site of the defect. Repeated loading and motion early during repair may disrupt the immature repair tissue and prevent healing. Successful repair replaces damaged tissue with new tissue consisting of a high concentration of fibroblasts and small blood vessels surrounded by a poorly organized matrix. After successful repair, the newly formed tissue begins to remodel. Cell density and vascularity decline, and excess tissue is resorbed. The collagen fibrils at the injury site assume a higher degree of orientation. Loading of the meniscus presumably influences remodeling of meniscal repair tissue, in the same manner that loading appears to affect collagen fibril organization in immature meniscal tissue.
Partial meniscal resection through the peripheral vascularized region or complete meniscal resection initiates production of repair tissue that can extend from the remaining peripheral tissue into the joint. Although the repair cells usually fail to replicate normal meniscal tissue, many authors have referred to this phenomenon as meniscal regeneration . Some repaired menisci grossly resemble normal menisci, but the functional capabilities and mechanical properties of this “regenerated” meniscal tissue have not been studied. Surgeons have reported meniscal regeneration in many clinical situations. Investigators have also examined the tissue produced by meniscal regeneration in animals. Meniscal regeneration can occur repeatedly in the same knee and occasionally occurs after total knee replacement. In rabbits, meniscal regeneration occurs more frequently on the medial side of the knee than on the lateral side, and development of degenerative changes in articular cartilage after a meniscectomy is inversely correlated with the extent of meniscal regeneration. Synovectomy appears to prevent meniscal regeneration, which suggests that synovial cells contribute to the formation of meniscal repair tissue. The mechanisms and the conditions that promote this type of repair, its functional importance, and the factors related to the predictability and frequency of meniscal regeneration remain unknown.
Repair in Avascular Portions of the Meniscus
The response of meniscal tissue to tears in the avascular portion resembles the response of articular cartilage to lacerations in many respects. Experimental studies show that a penetrating injury to the avascular region of the meniscus causes no apparent repair or inflammatory reaction. Meniscal cells in the injured region, like chondrocytes in the region of an injury limited to the articular cartilage, may proliferate and synthesize new matrix, but they appear to be incapable of migrating to the site of the defect or producing enough new matrix to fill it. The ineffective response of meniscal cells in the avascular region of the meniscus has led investigators to develop several methods to stimulate repair. Some promising approaches include creation of a vascular access channel to the injury site and stimulation of cell migration to the avascular region using implantation of a fibrin clot, an artificial matrix, or growth factors. Synovial abrasion has also been shown to stimulate proliferation of the synovial fringe into the meniscus and allows blood vessels to enter the avascular regions. Although early results appear promising, the quality of the repair tissue, its biomechanical properties, and the long-term results of these methods have not been evaluated.
Augmentation of Meniscus Healing
Given the well-established poor intrinsic healing potential of the meniscus, intense interest exists regarding methods to augment healing using cytokines, exogenous cells, and scaffolds. Fibroblast growth factor-2 and connective tissue growth factor have been evaluated in rabbit models, and vascular endothelial growth factor has been tested in a sheep meniscus tear model. Although these cytokines appear to have a positive effect on basic meniscal fibrochondrocyte biology, the challenge at this time is to identify the optimal carrier vehicles and dosage in order to translate these preclinical data to clinical trials. PRP, as a source of cytokines, may have some value in meniscus healing, but further study is required. Currently no human trials are being conducted to investigate the use of cytokines to improve meniscus healing.
Cell-based approaches have also been evaluated for augmentation of healing. Various sources of both autogenous and allogeneic cells have been evaluated using different carrier materials. Both differentiated cells such as chondrocytes and undifferentiated cells (mesenchymal stem cells) have been tested in animal models. Currently no human trials of cell-based therapies in meniscus healing are being conducted.
The use of scaffold materials to replace a portion of the damaged meniscus or to replace the entire structure is an appealing option and in theory can provide mechanical stability to the injured site while allowing for cell attachment and proliferation. A collagen-based scaffold (Collagen Meniscus Implant, Menaflex, ReGen Biologics, Glen Rock, NJ) and a resorbable porous polyurethane-based scaffold (Actifit, Orteq Sport Medicine, London, UK) have demonstrated satisfactory clinical outcome in up to 80% of cases at up to 10 years and 2 years of follow-up, respectively. Both of these devices are designed for partial meniscus replacement. Although it remains unclear whether the use of such scaffolds can affect the long-term sequelae of meniscectomy, early results are promising and may represent a new horizon in the treatment of these complex injuries. Further optimization of these materials may occur by incorporating undifferentiated cells into the scaffold.
Synovial joints allow the rapid controlled movements necessary to participate in sports. Normal function of these complex diarthrodial structures depends on the structural integrity and macromolecular composition of articular cartilage. Sports-related traumatic disruptions of cartilage structure or alterations in the macromolecular composition or organization change the biomechanical properties of the tissue, compromise joint function, and can lead to progressive pain and disability. Sports injuries to articular cartilage are more difficult to diagnose and treat compared with injuries to ligament, tendon, or bone because of the unique structure and function of articular cartilage.
The specialized composition and organization of hyaline articular cartilage make the diagnosis of many injuries difficult, but these characteristics also provide the unique biomechanical properties that permit normal synovial joint function. In the joint, cartilage distributes the loads of articulation, thereby minimizing peak stresses acting on the subchondral bone. The tensile strength of the tissue provides its structural integrity under such loads. Alterations in the mechanical properties of cartilage due to injury, disease, or increasing age have not been well defined, but the available information shows that these properties change with age and loss of structural integrity. Cartilage from skeletally immature joints (open growth plates) is much stiffer than cartilage from skeletally mature joints (closed growth plates). Older cartilage and fibrillated cartilage have much lower tensile stiffness and strength. Participation in sports often subjects the articular cartilage to intense repetitive, compressive high-energy impact forces that can cause tissue injury. These abnormally large forces generate high shear stresses at the cartilage-subchondral bone junction, causing matrix lesions and death of the articular chondrocytes that may lead to early osteoarthritis. Because cartilage lacks nerves, cartilage injuries do not cause pain and thus even when cartilage damage is suspected, it is difficult to diagnose. Less frequent but more severe sports injuries can acutely disrupt the articular surface by fracturing both cartilage and the underlying bone and may also lead to early osteoarthritis.
Composition of Articular Cartilage
Like the dense fibrous tissues and meniscus, articular cartilage consists of cells, matrix water, and a matrix macromolecular framework. Unlike the most dense fibrous tissues, cartilage lacks nerves, blood vessels, and a lymphatic system. The composition of articular cartilage is responsible for its unusual physiologic requirements, cell behavior, and responses to injury.
The chondrocyte is the only type of cell in cartilage. Chondrocytes contribute only 5% or less to the total volume of cartilage. Like other mesenchymal cells, chondrocytes surround themselves with their extracellular matrix and rarely form cell-to-cell contacts. In normal cartilage, they are isolated. Because the tissue lacks blood vessels, the cells depend on diffusion through the matrix for their nutrition and rely primarily on anaerobic metabolism.
Three distinct zones of chondrocytes are seen ( Fig. 1-4 ). The superficial tangential zone contains ellipsoidal cells with their long axes aligned parallel to the surface. The middle zone contains spherical cells that are randomly distributed throughout the region. The deep zone contains similar spherical cells that form columns aligned perpendicular to the tidemark and the calcified zone.