Bones, Joints, Tendons, and Ligaments

Bone at the Cellular Level

Structure and function can be discussed at the organ, tissue, and cellular levels. In this section, normal structure and function are briefly reviewed, beginning at the cellular level and including bone matrix and mineral. Cells directly involved with the structural integrity of bone include osteoblasts, osteocytes, and osteoclasts (Box 16-1).

Osteoblasts are cells on any bone surface (periosteal, endosteal, trabecular, intracortical) that produce bone matrix (osteoid), initiate the mineralization of this matrix (deposition of hydroxyapatite), and seemingly paradoxically initiate the resorption of this matrix by osteoclasts. Osteoblasts are derived from mesenchymal stem cells. Active osteoblasts are plump (Fig. 16-1), with abundant basophilic cytoplasm that is rich in rough endoplasmic reticulum and contains prominent Golgi apparatus and numerous mitochondria (Web Fig. 16-1). Inactive osteoblasts are disk-shaped with little cytoplasm because fewer organelles are needed for matrix synthesis and secretion (Fig. 16-2 and Web Fig. 16-2). Osteoblasts likely interact with osteocytes to assist in fine control of calcium homeostasis and detection of mechanical use and microscopic damage to bone, as discussed later. Indirect measurements of osteoblast activity are reflected in blood concentrations of the bone-specific isoform of alkaline phosphatase, an enzyme on the cell membrane of osteoblasts, and the noncollagenous protein, osteocalcin, that is secreted by osteoblasts and is present in bone matrix. Both are thought to have roles in mineralization and calcium ion homeostasis.

Osteocytes are osteoblasts that have been surrounded by mineralized bone matrix (Web Fig. 16-3) and are the most numerous and longest living of the bone cell types. They occupy small spaces in the bone called lacunae (singular: lacuna) and make contact with osteoblasts and other osteocytes by means of long cytoplasmic processes that pass through thin tunnels in the bone called canaliculi (singular: canaliculus). Under conditions of extreme stress to calcium homeostasis, osteocytes might have the ability to resorb perilacunar mineral and matrix, thus enlarging the lacuna (osteocytic osteolysis). This process apparently is rare and likely does not contribute significantly to development of osseous lesions. Osteocytes also retain a limited capacity to form bone and are exquisitely sensitive to mechanical strain in the form of shear stress. Other functions of osteocytes are somewhat speculative and are presented later in osteoblast-osteocyte interactions.

Osteoclasts are multinucleated cells that are derived from hematopoietic stem cells of the granulocyte-monocyte series and are responsible for bone resorption (Fig. 16-3). They have abundant eosinophilic cytoplasm and a specialized brush border along the margin of the cell that is adjacent to the bone surface that is being resorbed (Web Fig. 16-4). For osteoclasts to resorb bone, they must gain access to the bone surface that is usually covered by osteoblasts, and they must attach to the mineralized surface by transmembrane receptors in their sealing zones. These receptors bind to specific ligands in the matrix, likely residing in noncollagenous proteins. Osteoclasts are not able to bind to unmineralized bone matrix even though it contains these same ligands. Once bound to the matrix, the osteoclast resorbs bone in two stages. First, the mineral is dissolved by secretion of hydrogen ions through a proton pump located in the brush border. These hydrogen ions are derived from carbonic acid produced within the osteoclast from water and carbon dioxide by the enzyme carbonic anhydrase. Second, the collagen of the matrix is cleaved into polypeptide fragments by cysteine and metalloproteinases and cathepsins, particularly cathepsin K, released from the numerous lysosomes in the osteoclast and secreted through the brush border. The concavity in the bone created by the resorbed bone matrix is called a Howship’s lacuna, or a resorption lacuna. Physiologically, osteoclast activation is controlled by osteoblasts and bone marrow stromal cells (see later discussion of interactions). Calcitonin is a systemic inhibitor of osteoclasts. Osteoclasts have receptors for calcitonin and respond to this hormone by involuting their brush border and detaching from the bone surface. The activity of osteoclasts can be indirectly measured by determining serum concentrations of collagen degradation products (pyridinoline and deoxypyridinoline cross-links) or tartrate-resistant acid phosphatase (TRAP) activity. TRAP is a glycosylated monomeric metalloenzyme that is highly expressed by osteoclasts. TRAP staining also may be used as a cytochemical marker of osteoclasts in histologic sections (Web Fig. 16-5).

Osteoblast-osteocyte interactions are apparent from their connections to each other by thin, tortuous, cytoplasmic processes. This network of osteoblasts and osteocytes forms a functional membrane that separates the extracellular fluid bathing bone surfaces from the general extracellular fluid and can regulate the flow of calcium and phosphate ions to and from the bone fluid compartment (Fig. 16-4). Because of the large surface area of perilacunar and canalicular bone available for rapid ion exchange, significant amounts of calcium can be shifted from the bone fluid compartment to the extracellular fluid compartment without structural changes within the bone. In addition, this network allows osteocytes to detect alternations in the fluid flow within the bone extracellular fluid compartment. It is thought that such flow contributes to electric currents called streaming potentials. Changes in these streaming potentials caused by altered stress and strain on the bone or disruption of these potentials by microcracks (minuscule fractures within the bone visible only microscopically) might be detected by osteocytes with subsequent signaling to the overlying osteoblasts to initiate bone formation or resorption.

Osteoclast-osteoblast/stromal cell interactions are required for physiologic resorption of bone. The bone surface is protected from osteoclastic resorption by a continuous layer of osteoblasts and also by a very thin layer of unmineralized bone matrix normally present beneath resting osteoblasts (lamina limitans) (see Web Fig. 16-2). Osteoclasts are not able to bind to unmineralized bone matrix. For parathyroid hormone (PTH) to initiate bone resorption, receptors on the osteoblast bind PTH (osteoclasts do not usually express receptors for PTH). The binding of PTH to osteoblasts signals them to retract and secrete collagenases, which erode the unmineralized layer of matrix and allow osteoclasts access to a mineralized bone surface. In addition, osteoblasts and stromal cells that are activated by binding PTH and in response to a variety of other bone-resorbing stimuli (1, 25 dihydroxyvitamin D_3; interleukin [IL]-1, IL-6, and IL-11; tumor necrosis factor-α [TNF-α]; prostaglandin E₂ [PGE₂]; and glucocorticoids) express or secrete receptor activator for nuclear factor κ B ligand (RANKL), also called osteoclast differentiation factor (ODF). RANKL binds to the RANK receptor on osteoclasts and activates the resorption process. Osteoblasts and bone marrow stromal cells also can secrete osteoprotegerin (OPG), a RANK homolog that works by binding to RANKL, thus blocking the RANK-RANKL interaction and inhibiting the differentiation of osteoclast precursors into mature osteoclasts. OPG expression can be stimulated by transforming growth factor-β (TGF-β). Therefore osteoblasts and stromal cells have the ability to both up- and down-regulate osteoclastic bone resorption (Fig. 16-5). In conditions of inflammation and necrosis, inflammatory mediators, such as IL-1 and TNF-α, can stimulate osteoclasts directly, causing bone resorption independent of the presence of viable osteoblasts.

Bone at the Organic Matrix and Mineral Level

The mineralized matrix of bone provides the organ’s strength. Bone organic matrix consists of type I collagen and “ground substance” (the noncollagenous extracellular matrix: water, proteoglycans, glycosaminoglycans, noncollagenous proteins, and lipids, which are discussed later). Type I collagen polymers are secreted by osteoblasts and assembled into fibrils that are embedded in the ground substance and then mineralized. The type I collagen molecule is composed of three intertwined amino acid chains, unique to which is the hydroxylated form of the amino acid proline (hydroxyproline). Type I collagen molecules have extensive cross-linkages among the amino acid chains within the molecule and between adjacent molecules. Collagen molecules are deposited in rows with a gap between each molecule and with the rows staggered so that the molecules overlap by one-quarter of their length. This specific packing of the collagen molecules and the cross-linkages contribute to the strength and insolubility of the fibrous component of the bone matrix. Other than in rapidly deposited bone (woven bone found in primary trabeculae, bone of early fetal development, and pathologic conditions such as fracture repair), collagen fibers are arranged in parallel lamellae (singular: lamella) and the tissue is called lamellar bone. In cortical (compact) bone, lamellae are arranged concentrically (Fig. 16-6). In trabecular bone, the lamellae usually are arranged parallel with the bone surface. The collagen content of bone and its lamellar arrangement give bone its strength and flexibility. The ground substance of bone, which also is synthesized by osteoblasts, consists of noncollagenous proteins, proteoglycans, and lipids. Many of the noncollagenous proteins are cytokines that are capable of influencing bone cell activity. These cytokines, such as TGF-β, may play pivotal roles in controlling the extent of bone formation and resorption in normal remodeling and in disease (Fig. 16-7). Also, among the noncollagenous proteins are enzymes that can function in the degradation of collagen (e.g., matrix metalloproteinases) and can destroy inhibitors of mineralization (e.g., pyrophosphatases). Other noncollagenous proteins in the matrix can function as adhesion molecules and help bind cells to cells, cells to matrix, and mineral to matrix. Examples of these are osteonectin and osteocalcin. The role of proteoglycans in bone matrix is uncertain; however, there is some evidence that they influence bone cell differentiation and proliferative activity. Lipids may assist in binding calcium to cell membranes and in promoting calcification.

Bone mineral is in the form of a crystal called hydroxyapatite. Fully mineralized bone composes approximately 65% of the bone by weight and consists in part of calcium, phosphorus, carbonate, magnesium, sodium, manganese, zinc, copper, and fluoride. The mineral content gives bone its hardness. The production of osteoid (unmineralized organic matrix) by osteoblasts is followed by a period of maturation, after which mineral is deposited in exchange for water. Mineralization in woven bone is initiated within cytoplasmic blebs (matrix vesicles) of osteoblasts in the osteoid (see Web Fig. 16-1). Initiation of mineralization involves concentrating calcium, phosphorus, and other elements in these matrix vesicles to a level that causes precipitation of the mineral in the form of amorphous (not yet crystalline) hydroxyapatite. Matrix vesicles contain phospholipids and enzymes such as alkaline phosphatase and adenosine triphosphatase in their membranes. It is speculated that the membrane phospholipids attract calcium and phosphorus to the surface of the vesicle and that the alkaline phosphatase and adenosine triphosphatase enzymes might function in the pumping of these ions into the cell against a concentration gradient.

On reaching a critical mass, the amorphous mineral becomes crystalline. The crystalline hydroxyapatite pierces the lipid membrane of the matrix vesicle and extends to the gaps (holes) between collagen molecules. It is within these holes that the mineral crystals are first deposited in collagen. For the mineralization to spread beyond these gaps between the tropocollagen molecules, it is necessary for naturally occurring inhibitors of mineralization, such as inorganic pyrophosphate, in the matrix to be destroyed. Inorganic pyrophosphates are normal by-products of cellular metabolism and are deposited in unmineralized matrix by osteoblasts. The phosphatase enzymes described previously on the matrix vesicles have the ability to cleave these inorganic pyrophosphates and in doing so, destroy these inhibitors. Once the gaps are filled with mineral and the inhibitors of mineralization are destroyed, the process continues so that eventually the surfaces of collagen fibers, as well as spaces between collagen fibers, are mineralized. Initiation of mineralization in lamellar bone might not require matrix vesicles because glycoproteins, such as sialoprotein and osteonectin, can act as the nidus for the mineralization process.

Bone as a Tissue

In the cortex and subjacent to articular cartilage (subchondral bone), bone is organized into osteons (also called Haversian systems), which are cylinders of concentric layers of lamellae that are oriented parallel to the longitudinal axis of the bone and contain centrally located vessels and nerves (Fig. 16-8; see also Fig. 16-6). The bone between the osteons is called interstitial lamellae. Layers of bone oriented parallel to the internal and external circumference of the bone (beneath the endosteal and periosteal surfaces) are called circumferential lamellae. The osteonal system provides channels for the vascular supply to the cortex and also acts as tightly bound cables, giving the cortical bone strength and limited flexibility. This osteonal system also might be important in limiting propagation of microcracks in bone by diverting cracks along cement lines, which are collagen-poor, proteoglycan-rich seams between adjacent remodeling or modeling units (see later discussion). In H&E-stained histologic sections of decalcified bone, cement lines appear as basophilic lines.

In contrast to the dense compact bone of the cortex and the subchondral bone plates, the bone in the medullary cavity is in the form of anastomosing plates or rods and is called cancellous, trabecular, or spongy bone (see Figs. 16-8 and 16-10). The orientation of the trabeculae usually reflects adaptation (modeling) to mechanical stresses applied to the bone. This is readily apparent when examining the trabeculae in the femoral neck, which are oriented in lines and arcs that are perpendicular to the stress applied and are thicker and more numerous on the ventral aspect (side of compression) compared with the dorsal aspect (side of tension). The lamellae within a trabecula usually are arranged parallel to the surface of the trabecula and are not arranged into tubes or osteons, as they are in cortical bone.

In most species, bone undergoes a low but constant replacement called remodeling: old bone is resorbed and replaced by new bone. The basal level of this remodeling activity (number of sites in the skeleton being remodeled at any one time) is likely “programmed” for each species. The number of active remodeling sites, however, can be markedly increased or decreased in response to altered mechanical use (see later discussion). This turnover of old bone to new bone allows for the repair of accumulated microscopic injury in the bone (microfractures). In normal bone remodeling, slightly less bone is replaced than removed, leaving a small negative net change in bone mass with each remodeling cycle and explaining in part the reduced bone mass in aged animals. Furthermore, in disease states, such as hyperparathyroidism, resorption is often increased and formation decreased, leaving a significant net negative bone balance. Interestingly, in small, short-lived animals, such as the mouse and rat, cortical bone is not remodeled. In addition, not all areas of bone in larger species undergo remodeling.

The remodeling unit of cortical bone is called the osteon, whereas the remodeling unit of trabecular bone is called the basic structural unit. The shape of the osteon is cylindric; however, the basic structural unit has the contour of a shallow trench filled with parallel lamellae. The time required from initiation to completion of these remodeling units (from initiation of bone removal by osteoclasts to complete filling of the defect by osteoblasts), regardless of the type of bone, is estimated to be 3 to 4 months in humans. Basophilic cement lines that mark the limit of previous resorptive activity are usually somewhat scalloped, following the contours of the Howship’s lacunae created by osteoclasts, and are called reversal lines (where resorption stopped and the process was reversed by formation) (Fig. 16-9). Cement lines also can occur when osteoblast formation ceases and subsequently resumes, and these are called “resting lines” and are usually smooth, following the contour of the overlying surface.

The term modeling is used to describe change of the shape or contour of a bone in response to normal growth, altered mechanical use, or disease. In modeling, bone surfaces (periosteal, endosteal, intracortical, and trabecular) can go directly from resting to either formation or resorption, depending on the stimulus. This process allows the shape or size of bone to change, enables the medullary cavity to enlarge, and allows the overall shape of the bone to be maintained while it is growing. Modeling is in contrast to remodeling, in which resorption must precede formation to keep bone mass and shape constant.

Bone as an Organ

Individual bones of the skeleton vary in their manner of formation, growth, structure, and function. Flat bones of the skull develop by the process of intramembranous ossification, in which mesenchymal cells differentiate into osteoblasts and produce bone directly, in the absence of a preexisting cartilage model. In contrast, most bones develop from cartilaginous models by the process of endochondral ossification, in which cartilage is invaded by vessels, undergoes mineralization, and forms primary (diaphyseal) and secondary (epiphyseal) centers of ossification. Bones forming by endochondral ossification, such as the long bones of the appendicular skeleton and the vertebral bodies, are divided anatomically into epiphyses, metaphyseal growth plates (physes), metaphyses, and diaphyses (Fig. 16-10).

Blood Supply to Bone

Arterial blood from the systemic circulation enters bones through nutrient, metaphyseal, periosteal, and epiphyseal arteries (Fig. 16-11). Nutrient arteries penetrate the diaphyseal cortex through a nutrient foramen covered by strong, protective fascial attachments; once within the medulla, these arteries divide into proximal and distal intramedullary branches. Other arteries penetrating the cortex are the proximal and distal metaphyseal arteries, which are smaller and more numerous than the nutrient arteries. They penetrate the cortex and anastomose with the terminal branches of the nutrient arteries in the medullary cavity, protecting against infarction in case of obstruction of a nutrient artery.

Small periosteal arteries also pass through the diaphyseal cortex at sites of fascial attachment and can supply one quarter to one third of the outer cortex. The remainder of the cortex is supplied by the nutrient artery and its anastomotic branches. This blood flow is centrifugal (from medulla to periosteum) because of greater pressures in the intramedullary vessels. The chondrocytes of the physis nearest the epiphysis are supplied by epiphyseal arteries, whereas the chondrocytes of the physis nearest the metaphysis are supplied by branches of metaphyseal and nutrient arteries. As capillaries from these vessels approach the metaphyseal side of the physis, they make abrupt turns (loops), which are sites of predisposition to bacterial embolization in neonatal sepsis, sometimes resulting in osteomyelitis.

Bone Growth

Bone grows in length by interstitial growth within the metaphyseal growth plates (physes) (Fig. 16-12). The mineralized longitudinal septa of the growth plates serve as struts on which bone is deposited, a process called endochondral ossification. The metaphyseal growth plate is primarily responsible for lengthening the bone and is divided into a reserve or resting zone, a proliferative zone, and a hypertrophic zone (Fig. 16-13). The hypertrophic zone is sometimes further subdivided into zones of maturation, degeneration, and calcification. The resting or reserve zone serves as a source of cells for the proliferating zone in which cells multiply, accumulate glycogen, produce matrix, and become arranged in longitudinal columns. In the hypertrophic zone, chondrocytes secrete macromolecules that modify the matrix to allow capillary invasion and initiate matrix mineralization. The hypertrophic zone chondrocytes eventually die. The overall lengthening of the bone is due both to chondrocyte proliferation and hypertrophy. Calcification begins in the longitudinal septa of cartilaginous matrix between columns of chondrocytes. Matrix vesicles derived from chondrocytes (analogous to those described previously for mineralization of bone) form in the hypertrophic zone and initiate the mineralization process. The processes of mineralization and vascular invasion of the growth plate are codependent events. To supply salts for the mineralization, a nearby blood supply is necessary, and vascular invasion, a critical step in endochondral ossification, does not take place in mammalian growth plates unless there is mineralization of the longitudinal septum. Blood vessels from the metaphysis invade into the advancing growth plate, providing an entryway for osteoblasts that form bone on the cartilage spicules (see Fig. 16-13). The chondro-osseous junction in the metaphysis is a fragile lattice of bone-covered spicules of calcified cartilage (primary trabeculae). As the growth plate advances and elongates the metaphysis, the more mature trabeculae deeper in the metaphysis become fewer and thicker (secondary and tertiary trabeculae) and often contain residual fragments of cartilage.

Growth plates are thickest when growth is most rapid; as growth slows, the plate becomes thin and “closes” (becomes replaced by bone) at skeletal maturity. The age at which a growth plate closes varies, depending on site, species, and gender. For example, the physes of vertebrae usually remain open longer than the physes of the long bones. Androgens and estrogens play a major role in determining the time of growth plate closure, and early castration results in delayed growth plate closure with subsequent increased length of bones compared with intact individuals.

Growth of the epiphysis contributes both to the overall length of the bone and to the shape of the ends of the bone. This is accomplished by endochondral ossification at the articular-epiphyseal complex (AEC). The AEC is composed of permanent articular cartilage, as well as a subjacent temporary growth/epiphyseal cartilage that is vascularized and contains the same zones as the growth plate (Fig. 16-14). In the mature individual, endochondral ossification no longer occurs at the AEC. Although the articular cartilage remains throughout life, the epiphyseal cartilage is completely replaced by bone once growth has ceased (Fig. 16-15). Bone grows in width by intramembranous bone formation. Except for articular surfaces (including the ends of the vertebral bodies), the surfaces of bones are covered by periosteum. This covering is a thin membrane that is loosely attached to underlying bone except at heavy fascial attachments on bony prominences and at tendon insertions, where its attachments are strong and are associated with large vessels penetrating the underlying bone. Microscopically, the periosteum is composed of an outer fibrous layer that provides structural support and an inner osteogenic or cambium layer, capable of forming normal lamellar appositional bone on the cortex of growing bones and of forming abnormal woven bone formation in response to injury (Fig. 16-16). The periosteum is well supplied with lymph vessels and with fine myelinated and nonmyelinated nerve fibers that account for the intense pain that occurs with periosteal injury.

The periosteum covering the physis is called the perichondrial ring. The perichondrial ring adds new cartilage to the periphery of the physis, enabling it to expand in width as the animal grows. The metaphyseal cortex immediately adjacent to the perichondrial ring is normally very thin in the growing bone because its surfaces are the sites of very active osteoclastic bone resorption. Structurally, this area is the weakest part of the bone.

Articular Cartilage

Articular (hyaline) cartilage is normally a white to blue-white material with a smooth, moist surface. Cartilage thickness is greatest in the young and at sites of maximal weight bearing. Thinning and yellow discoloration occur in old age. At its margins, articular cartilage merges with a periosteal surface that is lined by fibrous tissue contiguous with the synovial membrane. Synovial fossae are normal depressions on non–weight-bearing articular cartilage surfaces (Fig. 16-17) that develop bilaterally in the larger appendicular joints of the horse, pig, and ruminant. These fossae are not present at birth but are fully formed by skeletal maturity. The function of synovial fossae is not known; however, they are significant because they are often mistaken for lesions. Adult articular cartilage contains no nerves or blood or lymph vessels, and its nutrients are obtained by diffusion from synovial fluid and to a lesser extent from subchondral vessels. In the immature skeleton, articular cartilage overlies the temporary growth cartilage of the epiphysis (epiphyseal cartilage). The epiphyseal cartilage is highly dependent on a network of blood vessels that originate from the perichondrium and the subchondral bone. Similar to the physis, it undergoes endochondral ossification and thereby contributes to the growth/development of the epiphysis. At skeletal maturity, the epiphyseal cartilage has been entirely replaced by bone, and the only joint cartilage remaining is the adult articular cartilage, which includes a thin subjacent zone of mineralized cartilage. In skeletally mature individuals, the junction between the unmineralized articular cartilage and the deeper mineralized cartilage forms a thin basophilic line in H&E stained sections called the tidemark (see Fig. 16-15). As animals age, multiple tidemarks can be formed, indicating an advance (thickening) of the mineralized layer of articular cartilage and thinning of the overlying nonmineralized articular cartilage. The mineralized cartilage serves to anchor articular cartilage to subchondral bone and limits the diffusion of substances between bone and cartilage.

Articular cartilage is 70% to 80% water by weight. It is a viscoelastic, hydrated fiber-reinforced gel that contains chondrocytes, collagen fibers (mostly type II), noncollagenous proteins, and proteoglycan aggregates. The largest proteoglycan in articular cartilage is aggrecan, which is composed of the immensely long, nonpolysulfated glycosaminoglycan, hyaluronan, to which core proteins are attached in a perpendicular arrangement like bristles on a brush. Similarly, smaller, negatively charged polysulfated glycosaminoglycans (chondroitin sulfate and keratan sulfate) are attached to the core proteins. The large content of negatively charged polysaccharide chains in the aggrecan aggregate is responsible for the extremely high osmotic swelling pressure of cartilage. The swelling pressure provided by aggrecan is counteracted by the resistance of the intact type II collagen fibers, giving cartilage its characteristic properties of being able to resist compressive forces and having a high tensile strength. Collagen fibers are arranged in arcades so that the tops of the arcades are parallel to the articular surface, and the sides are perpendicular to the surface and parallel with the radial or intermediate zone of chondrocytes. Functionally, the superficial zone of articular cartilage resists shearing forces, the middle zone functions in shock absorption, and the mineralized deep zone of cartilage serves to attach articular cartilage to the subchondral bone by its irregular (and therefore interlocking) interfaces. By scanning electron microscopy, the surface of articular cartilage is not smooth, but rather contains numerous depressions that can serve as reservoirs for synovial fluid.

Articular cartilage contains a single population of cells called chondrocytes that are responsible for the production, maintenance, and turnover of intercellular substances. In routine histologic sections, chondrocytes appear to be present within lacunae. However, although the lacunae within which osteocytes reside contain fluid spaces, the apparent lacunae of chondrocytes have been clearly demonstrated to be the result of shrinkage artifact. In fact, the chondrocyte cell membrane is in direct contact with the pericellular matrix and contains surface receptors for matrix components (e.g., hyaluronan).

Normal matrix turnover is enzymatic and this process is balanced by enzyme inhibitors. Proteases capable of degrading aggrecans are called aggrecanases and are members of the ADAM (A Disintegrin And Metalloprotease) protein family, whereas proteases capable of breaking the peptide bonds in collagen are called collagenases. Disease occurs if there is increased destruction or decreased synthesis of matrix components. It is important to remember that compared with bone that normally renews itself by remodeling, articular cartilage has extremely poor regenerative capabilities. Although there is evidence for proteoglycan synthesis in normal articular cartilage, the turnover of cells and type II collagen occurs at an extremely low rate. Correspondingly, the cellularity of articular cartilage declines with age.

Articular Capsule/Synovium/Synovial Fluid

The articular capsule consists of outer fibrous and inner synovial tissue layers. The outer layer is a heavy sheath that contributes to joint stability and at its insertion, attaches to bone at the margins of the joint, thereby enclosing a segment of bone of variable length within the joint cavity. It is well supplied with blood vessels and nerve endings. The inner synovial tissue layer is called the synovial membrane and covers all the inner surfaces of the joints except for the surface of the articular cartilage. The synovial membrane is normally very thin, lacks a basement membrane, and is barely visible grossly. In fact, collection of normal synovial membrane for histologic evaluation is best done immediately after opening the joint because otherwise, it will retract and become difficult to locate. The inner surface of the synovial membrane may be flat or may contain tiny projections (villi). Synovial intimal or lining cells, one to four cells thick, form a discontinuous surface layer and include two cell types: (1) A cells, which are macrophages that are responsible for the removal of microbes and the debris resulting from normal wear and tear in the joint (Fig. 16-18) and (2) B cells, which are fibroblast-like cells that produce synovial fluid. Normal synovial fluid contains hyaluronic acid, lubricin (a water-soluble glycoprotein), proteinases, and collagenase and is clear, colorless to pale yellow, and viscous. In addition to lubricating the joint surfaces, it supplies oxygen and nutrients to and removes carbon dioxide and metabolic wastes from the chondrocytes in articular cartilage. The synovial subintima can be classified according to the type of tissue that predominates (areolar, adipose, or fibrous), and contains blood and lymph vessels that supply and drain the intraarticular structures. Adipose tissue sometimes accumulates in the deep layers of the synovium, forming fat pads that serve as soft cushions in joint cavities (e.g., the infrapatellar fat body of the stifle joint).

Joint lubrication depends on the microscopic roughness, elasticity, and hydration of articular cartilage and on the presence of hyaluronic acid and lubricin in synovial fluid. The lubricating properties of lubricin depend on its ability to bind to articular cartilage, where it retains a protective layer of water molecules. In contrast, hyaluronic acid, the molecule that makes synovial fluid viscous, has largely been excluded as a lubricant of the cartilage-on-cartilage bearing and instead, lubricates the site of surface contact between synovium and cartilage. When pressure is applied to the joint, such as during weight bearing, the articular surface is supplied with pressurized fluid that carries most of the load. This fluid is synovial fluid admixed with water that is released from the underlying cartilage, subsequent to pressure of weight bearing. When the load is removed, the water returns to the cartilage because of the hydrophilic properties of proteoglycans. This flushing of fluid in and out of the articular cartilage enables nutrients to enter cartilage from the synovial fluid and waste products to be removed.

Subchondral Bone

The subchondral bone acts to support the overlying cartilage and dissipate concussive forces to the peripheral cortical bone (Fig. 16-19). The thickness of the subchondral bone plate varies in proportion to the degree of weight bearing. Increased subchondral bone thickness is a common pathological finding in osteoarthritis. In larger animals (nonrodent species), subchondral bone often is composed of compact osteonal bone rather than trabecular bone.

Articular Cartilage

Although articular cartilage contains metabolically active cells, it has a limited response to injury and minimal capacity for repair that is a result in large part to its lack of a blood supply. Superficial cartilage defects (cartilage erosions) that do not penetrate into subchondral bone persist for long periods with few or no histologic changes. Clusters or clones of chondrocytes (evidence of local chondrocyte replication in response to injury) may be present, particularly along the margins of the defect, but are ineffective in filling it. Progression of the lesion, with matrix degeneration and eventual loss of the remaining articular cartilage, may occur over time, particularly if thickened (sclerotic) subchondral bone is present subjacent to the defect. In contrast, if a cartilaginous defect extends into subchondral bone (cartilage ulceration), allowing mesenchymal cells access to the defect, it is quickly filled with vascular fibrous tissue that often undergoes metaplasia to fibrocartilage but rarely, if ever, to hyaline cartilage. Formation of fibrocartilage can be hastened in full-thickness cartilaginous defects by exercise or prolonged passive motion. Injury to articular cartilage is not painful unless the synovium or subchondral bone is involved. Having no blood supply of its own, articular cartilage does not participate directly in the inflammatory response but is very much affected by inflammation in the synovium, subchondral bone, or subarticular growth (vascularized) cartilage of the epiphysis in young animals. Given that alternating compression and release of normal weight bearing facilitates the diffusion of fluid with nutrients into the articular cartilage and fluid with metabolic waste products out of articular cartilage, it follows that constant compression or lack of weight bearing leads to atrophy (thinning) of articular cartilage.

Sterile injury to cartilage can be a consequence of trauma, joint instability, or lubrication failure because of changes in synovial fluid, synovial membrane, or incongruity in joint surfaces. Destruction of articular cartilage in response to sterile injury and infectious inflammation is mediated by a combination of enzymatic digestion of matrix and failure of matrix production when chondrocytes become degenerate or necrotic. These changes can be initiated by damage to the cartilage directly or may occur indirectly, secondary to lesions in the synovium.

Matrix metalloproteinases are enzymes capable of matrix digestion; they are normal constituents of the matrix, but are present in an inactive form. Matrix metalloproteinases can be broadly categorized as gelatinases, collagenases, and stromelysins. Collagenases are most capable of digestion of collagen fibers; gelatinases digest type I collagen and basement membrane collagens but are less effective against type II collagen of cartilage. Stromelysins destroy noncollagenous proteins. Matrix metalloproteinases can be activated by products of degenerating or reactive chondrocytes and inflammatory cells. In addition, tissue inhibitors of metalloproteinases (TIMPs) are present in the matrix, acting as a control on the destructive effects of activated metalloproteinases. As mentioned earlier, proteases capable of degrading aggrecans are called aggrecanases and are members of the ADAM protein family. The loss of proteoglycans from cartilage alters the hydraulic permeability of the cartilage, thereby interfering with joint lubrication and leading to further mechanically induced injury to the cartilage. The loss of proteoglycans, with subsequent inadequate lubrication of the articular surface, leads to disruption of collagen fibers on the surface of articular cartilage. Grossly, the surfaces of affected areas of cartilage are yellow-brown and have a dull, slightly roughened appearance. As more proteoglycans are lost, the collagen fibers condense and fray (fibrillation) with multiple clefts and/or fissures forming along the vertical axis of the arcades of collagen fibers (Figs. 16-27 and 16-28). The vertical axis of the collagen fibers in these arcades is perpendicular to the plane of movement of the joint. Fibrillation is accompanied by loss of surface cartilage (erosion) and eventual thinning of the articular cartilage. Subsequent to fibrillation and erosion, necrosis of individual chondrocytes occurs in the radial zone, resulting in hypocellularity of the remaining cartilage. In response to the fibrillation, erosion, and necrosis of chondrocytes, remaining chondrocytes can undergo regenerative hyperplasia (cluster or clone formation), but the ability of chondrocytes in the adult to divide is limited and the regenerative attempt is almost always ineffective. Loss of articular cartilage can become complete, resulting in exposure of subchondral bone (ulceration), which typically is eburnated (thickened/sclerotic) (Fig. 16-29).

Articular Capsule/Synovium/Synovial Fluid

Intraarticular prostaglandins, nitric oxide, TNF-α, IL-1, and neurotransmitters—such as substance P, among other cytokines and chemokines—are increased in degenerative and inflammatory joint disease. Prostaglandins and nitric oxide inhibit proteoglycan synthesis in synovium and chondrocytes; this reduction in proteoglycan content can lead to degeneration and loss of the cartilage (see previous discussion). IL-1 and TNF-α are cytokines secreted by activated macrophages (synovial type A cells or subintimal macrophages); they promote secretion of prostaglandins, nitric oxide, and neutral proteases from synovial fibroblasts and chondrocytes. Increasing concentrations of these agents decreases matrix synthesis and increases matrix destruction. Cytokines and growth factors that can be anabolic to cartilage include IL-6, TGF-β, and insulin-like growth factor (IGF).

Lysosomal enzymes (collagenase, cathepsins, elastase, and arylsulfatase) and neutral proteases, which are capable of degrading proteoglycans or collagen, can be derived from inflammatory cells, synovial lining cells, and chondrocytes.

The synovial membrane commonly responds to injury by villous hypertrophy and hyperplasia (Fig. 16-30), hypertrophy and hyperplasia of synoviocytes, and pannus formation (see later discussion). Villous hypertrophy occurs with or without synovitis. The proportions of A (macrophages) and B (fibroblast-like) cells in the synovium also can change in various disease processes. Fragments of articular cartilage can adhere to the synovium, where they are surrounded by macrophages and giant cells. Larger pieces of detached cartilage (as in osteochondrosis) can float free and survive as chondral or osteochondral fragments (“joint mice”) that continue to remain viable through nutrient supply from synovial fluid.

Inflammatory cell infiltrates (see the discussion on infectious and noninfectious arthritis in the section on Inflammatory Lesions) in the synovial membrane can impair fluid drainage from the joint and can cause joint fluid to lose some of its lubricating properties as the result of degradation of hyaluronic acid by the superoxide-generating systems of neutrophils.

Pannus can develop in association with chronic infectious fibrinous synovitis and with some immune-mediated diseases; the classic example is rheumatoid arthritis. Pannus is a fibrovascular and histiocytic tissue (also called an inflammatory granulation tissue) that arises from the synovial membrane and spreads as a membrane over articular cartilage (Figs. 16-31 and 16-32). In the pannus, tissue histiocytes and monocytes of bone marrow origin transform into macrophages and they, along with the collagenases from fibroblasts, cause lysis and destruction of the underlying cartilage (Fig. 16-33). In time, if both opposing cartilaginous surfaces are involved, the fibrous tissue can unite the surfaces, causing fibrous ankylosis (fusion of the joint). In some cases of immune-mediated arthritis, pannus is present in the subchondral bone marrow (bone marrow pannus), as well as in the synovium, and may penetrate into the overlying articular cartilage.

Sterile degenerative changes in articular cartilage are often accompanied by the formation of periarticular osteophytes (Fig. 16-34) and by some degree of secondary synovial inflammation and hyperplasia. The synovitis is characterized by the presence of variable numbers of plasma cells, lymphocytes, and macrophages in the synovial subintima (beneath the layers of synoviocytes) and by hyperplasia and hypertrophy of synovial lining cells. The pathogenesis of this synovitis is not known but is suspected to be the result in part of the presence of degenerate cartilage debris within the joint. The pathogenesis of osteophytes also is unclear. They can arise from mesenchymal cells with chondro-osseous potential within the synovial membrane at the junction of the synovial membrane with the perichondrium/periosteum, just peripheral to the articular cartilage, or on the surface of the bone where the articular capsule and the periosteum merge. Osteophytes do not grow continuously, but once formed, they persist as multiple periarticular spurs of bone. These spurs can be confined within the joint cavity if they arise from the perichondrium or can be protrusions from the periosteal surface of the bone if they arise from the insertion site of the joint capsule with the periosteum. Osteophytes can result from mechanical instability within the joint, causing stretching or tearing of the insertions of the articular capsule or ligaments, or they can form from stimulation by cytokines such as TGF-β that are released from reactive or degenerating mesenchymal cells within the joint.

Because of their antiinflammatory effects, glucocorticoids often are therapeutically injected into joints. Although the usual result is decreased pain and inflammation, glucocorticoid injection sometimes is followed by a rapid progression of degenerative changes within the joint that is designated “steroid arthropathy.” These degenerative changes relate to the antianabolic effects of glucocorticoids on chondrocytes, in which synthesis of cartilaginous matrix is reduced, proteoglycans are depleted, repair is retarded, and the mechanical strength of cartilage is reduced.

Subchondral Bone

With the loss of matrix proteoglycans in degenerate articular cartilage, a greater component of concussive forces is transmitted to the subchondral bone. The bone responds to the increased mechanical use by decreasing resorption and increasing formation, resulting in a net increase in amount of bone per unit area (increased subchondral bone density or subchondral sclerosis). If the cartilage ulcerates to the level of the bone and the joint is still being used, the surface of the dense subchondral bone can become smooth and shiny (eburnation) (see Fig. 16-29, A). There is interest in the possible role subchondral bone might play in initiating cartilage damage in degenerative joint disease, as some studies have reported an increase in subchondral bone thickness/density before the development of articular cartilage lesions. Some believe that this dense bone is less effective in dissipating normal concussive forces and causes some of the impact to be deflected back into the articular cartilage, causing injury to the chondrocytes.

Abnormalities of Growth and Development

Metabolic Bone Diseases

Metabolic bone diseases are systemic skeletal diseases that are generally of nutritional, endocrine, or toxic origin (Table 16-1 and Web Table 16-1) and occur in both growing and adult skeletons. Metabolic bone diseases are often called osteodystrophies, which is a general term that implies defective bone formation. The classic metabolic osteodystrophies are osteoporosis, fibrous osteodystrophy, rickets, and osteomalacia. These terms imply specific pathologic changes but do not necessarily imply specific causes. For example, osteoporosis can be the result of a calcium deficiency, glucocorticoid therapy, or physical inactivity. In addition, different osteodystrophies can coexist in the same skeleton. For example, the skeleton of an animal with a severe calcium deficiency accompanied by excess dietary phosphorus might have features of both osteoporosis and fibrous osteodystrophy. In practice, most nutritional deficiencies in domestic animals do not involve a single element. More often, deficiencies are multiple, relatively mild, and do not produce the “classic” lesions that occur under experimental conditions.