Biomechanics & Biomaterials: Introduction
Authors: Ranjan Gupta, MD, Vincent Caiozzo, PhD, Stephen D. Cook, PhD, Robert L. Barrack, MD, & Harry B. Skinner, MD, PhD
Orthopedic surgery is the branch of medicine concerned with restoring and preserving the normal function of the musculoskeletal system. As such, it focuses on bones, joints, tendons, ligaments, muscles, and specialized tissues such as the intervertebral disk. Over the last half century, surgeons and investigators in the field of orthopedics have increasingly recognized the importance that engineering principles play both in understanding the normal behavior of musculoskeletal tissues and in designing implant systems to model the function of these tissues. The goals of the first portion of this chapter are to describe the biologic organization of the musculoskeletal tissues, examine the mechanical properties of the tissues in light of their biologic composition, and explore the material and design concepts required to fabricate implant systems with mechanical and biologic properties that will provide adequate function and longevity. The subject of the second portion of the chapter is gait analysis.
BASIC CONCEPTS & DEFINITIONS
Most biologic tissues are either porous materials or composite materials. A material such as bone has mechanical properties that are influenced markedly by the degree of porosity, defined as the degree of the material's volume that consists of void. For instance, the compressive strength of osteoporotic bone, which has increased porosity, is markedly decreased in comparison with the compressive strength of normal bone. Like composite materials, alloys consist of two or more different metallic elements that are in solution. Although composite materials can be physically or mechanically separated, alloyed materials cannot.
Generally, composites are made up of a matrix material, which absorbs energy and protects fibers from brittle failure, and a fiber, which strengthens and stiffens the matrix. The performance of the two materials together is superior to that of either material alone in terms of mechanical properties (eg, strength and elastic modulus) and other properties (eg, corrosion resistance). The mechanical properties of various types of composite materials differ, based on the percentage of each substance in the material and on the principal orientation of the fiber. The substances in combination, however, are always stronger for their weight than is either substance alone. Microscopically, bone is a composite material consisting of hydroxyapatite crystals and an organic matrix that contains collagen (the fibers).
The mechanical characteristics of a material are commonly described in terms of stress and strain. Stress is the force that a material is subjected to per unit of original area, and strain is the amount of deformation the material experiences per unit of original length in response to stress. These characteristics can be adequately described by a stress–strain curve (Figure 1–1), which plots the effect of a uniaxial stress on a simple test specimen made from a given material. Changes in the geometric dimensions of the material (eg, changes in the material's area or length) have no effect on the stress–strain curve for that material.
Mechanical characteristics can also be illustrated by a load-elongation curve, in which the slope of the initial linear portion depicts the stiffness of a given material. Although similar in appearance to the stress–strain curve, the load-elongation curve for a given material can be altered by changes in the material's diameter (cross-sectional area) or length. For instance, doubling the cross-sectional area of a test specimen while maintaining the original length will double the stiffness because the increased diameter doubles the load to failure (ie, it doubles the force that a material can withstand in a single application) without changing the total elongation. Conversely, doubling the length of the test specimen while maintaining the original diameter will decrease the stiffness by half because doubling the length in turn doubles the elongation without changing the load to failure.
Because of this difference between the stress–strain curve and load-elongation curve, any comparison of the characteristics of specimens requires that the same type of curve be used in the evaluation. If the load-elongation curve is used, the geometric dimensions of the specimens must also be the same. In this chapter, subsequent discussions pertain to the stress–strain curve, although differing terminology in the load-elongation curve is noted parenthetically.
The initial linear or elastic portion of the stress–strain curve (see Figure 1–1) depicts the amount of stress a material can withstand before permanently deforming. The slope of this line is termed the modulus of elasticity (stiffness) of the material. A high modulus of elasticity indicates that the material is difficult to deform, whereas a low modulus indicates that the material is more pliable. The modulus of elasticity is an excellent basis on which different materials can be compared. When materials such as those used in implants are compared, however, it is important to remember that the modulus of elasticity is a property only of the material itself and not of the structure. Implant stiffness in bending—or, more correctly, flexural rigidity—is a function both of material elastic modulus and of design geometry.
The proportional limit, or p, of a material is the stress at which permanent or plastic deformation begins. The proportional limit, however, is difficult to measure accurately for some materials. Therefore, a 0.2% strain offset line parallel to the linear region of the curve is constructed, as shown in Figure 1–1. The stress corresponding to this line is defined as the yield stress, or y. If stress is removed after the initiation of plastic deformation (point A in Figure 1–1), only the elastic deformation denoted by the linear portion of the stress–strain curve is recovered. The ultimate tensile strength (failure load), or u, is the maximal stress that a material can withstand in a single application before it fails.
When subjected to repeated loading in a physiologic environment, a material may fail at stresses well below the ultimate tensile strength. The fatigue curve, or S-N curve, demonstrates the behavior of a metal during cyclic loading and is shown in Figure 1–2. Generally, as the number of cycles (N) increases, the amount of applied stress (S) that the metal can withstand before failure decreases. The endurance limit of a material is the maximal stress below which fatigue failure will never occur regardless of the number of cycles. Fatigue failure will occur if the combination of local peak stresses and number of loading cycles at that stress are excessive. Although most materials exhibit a lower stress at failure with cyclic loading, some do not, such as pyrolytic carbon, making it appropriate for high-cycle applications such as heart valves. Environmental conditions strongly influence fatigue behavior. The physiologic environment, which is corrosive, can significantly reduce the number of cycles to failure and the endurance limit of a material.
Materials can be evaluated in terms of ductility, toughness, viscoelasticity, friction, lubrication, and wear. These properties are introduced here, and many of them are explored in detail in subsequent sections.
Ductility is defined as the amount of deformation that a material undergoes before failure and is characterized in terms of total strain. A brittle material will fail with minimal strain caused by propagation of a crack because the yield stress is higher than the tensile stress. A ductile material, however, will fail only after markedly increased strain and decreased cross-sectional area. Polymethylmethacrylate (PMMA, a polymer) and ceramics are brittle materials, whereas metals exhibit relatively more ductility. Environmental conditions, especially changes in temperature, can alter the ductility of materials.
Toughness is defined as the energy imparted to a material to cause it to fracture and is measured by the total area under the stress–strain curve.
Because all biologic tissues are viscoelastic in nature, a thorough understanding of viscoelasticity is essential. A viscoelastic material is one that exhibits different properties when loaded at different strain rates. Thus, its mechanical properties are time-dependent. Bone, for example, absorbs more energy at fast loading rates, such as in high-speed motor vehicle accidents, than at slow loading rates, such as in recreational snow skiing.
Viscoelastic materials have three important properties: hysteresis, creep, and stress relaxation. When a viscoelastic material is subjected to cyclic loading, the stress–strain relationship during the loading process differs from that during the unloading process (Figure 1–3). This difference in stress–strain response is termed hysteresis. The deviation between loading and unloading processes depends on the degree of viscous behavior. The area between the two curves is a measure of the energy lost by internal friction during the loading process. Creep, which is also called cold flow and observed in polyethylene components, is defined as a deformation that occurs in a material under constant stress. Some deformation is permanent, persisting even when the stress is released. The decrease in stress associated with a constant strain over time is a result of stress relaxation, a phenomenon evident, for example, in the loosening of fracture fixation plates. The time necessary to attain creep, or stress relaxation equilibrium, is an inherent property of the material.
Friction refers to the resistance between two bodies when one slides over the other. Friction is greatest at slow rates and decreases with faster rates. This is because the surface asperities (peaks) tend to adhere to one another more strongly at slower rates. Mechanisms of lubrication reduce the friction between two surfaces. Several lubrication mechanisms are present in articular cartilage to overcome friction processes in normal joint motion. Similarly, mechanisms are present in polyethylene-metal articulations to overcome friction in joint replacements.
Wear can occur from friction and is defined as the removal of surface material by mechanical motion. Wear is always observed between two moving surfaces, but lubrication mechanisms act to reduce the detrimental effects of excessive wear. Three types of wear mechanisms are apparent in normal and prosthetic joint motion: abrasive, adhesive, and three-body wear. Abrasive wear is the generation of material particles from a softer surface when it moves against a rougher, harder surface. An example of the product of abrasive wear is sawdust, which results from the movement of sandpaper against a wood surface. The amount of wear depends on factors such as contact stress, hardness, and finish of the bearing surfaces.
Adhesive wear results when a thin film of material is transferred from one bearing surface to the other. In prosthetic joints, the transfer film can be either polyethylene or the passivated (corrosion-resistant) layer of metal. Regardless of the material, wear occurs in the surface that loses the transfer film. If the particles from the transfer film are shed from the other surface as well, they behave as a third body and also result in wear.
Three-body wear occurs when another particle is located between two bearing surfaces. Cement particles act as third bodies in prosthetic joints. Implant designers continue to search for compatible substances that reduce friction at articulating surfaces and thereby reduce the amount of wear debris generated. Wear of polyethylene is the dominant problem in total joint replacement today because the wear debris generated is biologically active and leads to osteolysis.
BIOMECHANICS IN ORTHOPEDICS
An analysis of the factors that influence normal and prosthetic joint function requires an understanding of free-body diagrams as well as the concepts of force, moment, and equilibrium.
Force, Moment, & Equilibrium
Forces and moments are vector quantities—that is, they are described by point of application, magnitude, and direction. A force represents the action of one body on another. The action may be applied directly (eg, via a push or a pull) or from a distance (eg, via gravity). A normal tensile or compressive force is applied perpendicular to a surface, whereas a shear force is applied parallel to a surface. A force that is applied eccentrically produces a moment.
The force generated by gravity on an object acts at the center of gravity. An object that is symmetric has its center of gravity in the geometrically centered position, whereas an object that is asymmetric has its center of gravity closer to its "heavier" end. The center of gravity for the human body is the resultant of the individual centers of gravity from each segment of the body. Therefore, as the body segments move, the center of gravity changes accordingly and may even lie outside the body in extreme positions, such as encountered in gymnastics. A moment is defined as the product of the quantity of force and the perpendicular distance between the line of action of the force and the center of rotation. A moment usually results in a rotation of the object about a fixed axis.
Newton's first law states that a body (or object) is in equilibrium if the sum of the forces and moments acting on the body are balanced; therefore, the sum of forces and moments for each direction must equal zero. The concept of equilibrium is important in understanding and determining force–body interactions, such as the increased joint reaction force occurring in an extended arm because of an external weight and such as the increased joint reaction force occurring in the hip at a specific moment during walking.
A free-body diagram can be used schematically to represent all the forces and moments acting on a joint. The concepts of equilibrium can be extended to determine joint reaction or muscle forces for different conditions, as demonstrated in the following two examples.
Example 1: Determine the force on the abductor muscle of a person's hip joint (the abductor force, or FAB) and the joint reaction force (the FJ) when the person is standing on one leg. The weight of the trunk, both arms, and one leg is 5/6 of the total weight (w) of the person. As illustrated in Figure 1–4, this weight will tend to rotate the body about the femoral head and is counteracted by the pull of the abductor muscles on the pelvis. The necessary equation to solve for FAB is as follows:
In solving the equation, assume that a = 5 cm and that b = 15 cm.
After this equation is solved, two of the three forces are known. The remaining force (the FJ) can be determined from a force triangle (see Figure 1–4) because, according to Newton's first law, the sum of forces must equal zero.
Example 2: Determine the force on a person's deltoid muscle (the deltoid force, or FD) and the force of the joint acting about the shoulder (FJ) when the person holds a metal weight (w) at arm's length (Figure 1–5). The weight of the arm is ignored because only the increase in forces about the shoulder caused by the metal weight is to be determined. FD is determined by summing the moments about the joint center. The necessary equation is as follows:
In solving the equation, assume that a = 5 cm and that b = 60 cm.
After this equation is solved, an FJ of 1150 N is determined using a force triangle (see Figure 1–5).
Moments of Inertia
The orientation of the bone's or implant's cross-sectional area with respect to the applied principal load also greatly influences the biomechanical performance. Bending and torsion occur in long bones and are important considerations in the design of implants. In general, the farther that material mass is distributed from the axis of bending or torsion while still retaining structural integrity, the more resistant the structure will be to bending or torsion. The area moment of inertia is a mathematical expression for resistance to bending, and the polar moment of inertia is a mathematical expression for resistance to torsion. Both types of moment of inertia relate the cross-sectional geometry and orientation of the object with respect to the applied axial load. The larger the area moment of inertia or the polar moment of inertia, the less likely the material will fail. Figure 1–6 summarizes the area moments of inertia for representative shapes important to orthopedic surgery. Creating an open slot in an object will significantly decrease the polar moment of inertia of the object.
Knowledge of moments of inertia is important for understanding mechanical behavior in relation to object geometry. For instance, the length of the long bones predisposes them to high bending moments. Their tubular shape helps them resist bending in all directions, however. This resistance to bending is attributable to the large area moment of inertia because the majority of bone tissue is distributed away from the neutral axis. The concept of moment of inertia is crucial in the design of implants that are exposed to excessive bending and torsional stresses.
BIOLOGIC TISSUES IN ORTHOPEDICS
The functions of the musculoskeletal system are to provide support for the body, to protect the vital organs, and to facilitate easy movement of joints. The bone, articular cartilage, tendon, ligament, nerve, and muscle all interact to fulfill these functions. The musculoskeletal tissues are integrally specialized to perform their duties and have excellent regenerative and reparative processes. They also adapt and undergo compositional changes in response to increased or decreased stress states. Specialized components of the musculoskeletal system, such as the intervertebral disk, are particularly suited for supporting large stress loads while resisting movement.
Bones are dynamic tissues that serve a variety of functions and have the ability to remodel to changes in internal and external stimuli. Bones provide support for the trunk and extremities, provide attachment to ligaments and tendons, protect vital organs, and act as a mineral and iron reservoir for the maintenance of homeostasis.
Bone is a composite consisting of two types of material. The first material is an organic extracellular matrix that contains collagen, accounts for approximately 30–35% of the dry weight of bone, and is responsible for providing flexibility and resilience to the bone. The second material consists primarily of calcium and phosphorous salts, especially hydroxyapatite [Ca10(PO4)6(OH)2], accounts for approximately 65–70% of the dry weight of bone, and contributes to the hardness and rigidity of the bone. Microscopically, bone can be classified as either woven or lamellar.
Woven bone, which is also called primary bone, is characterized by a random arrangement of cells and collagen. Because of its relatively disoriented composition, woven bone demonstrates isotropic mechanical characteristics, with similar properties observed regardless of the direction of applied stress. Woven bone is associated with periods of rapid formation, such as the initial stages of fracture repair or biologic implant fixation. Woven bone, which has a low mineral content, remodels to lamellar bone.
Lamellar bone is a slower forming, mature bone that is characterized by an orderly cellular distribution and regular orientation of collagen fibers (Figure 1–7). The lamellae can be parallel to one another or concentrically organized around a vascular canal called a Haversian system or osteon. At the periphery of each osteon is a cement line, a narrow area containing ground substance primarily composed of glycosaminoglycans. Neither the canaliculi nor the collagen fibers cross the cement line. Biomechanically, the cement line is the weakest link in the microstructure of bone. The organized structure of lamellar bone makes it anisotropic, as seen in the fact that it is stronger during axial loading than it is during transverse, or shear, loading.
Bone can be classified macroscopically as cortical tissue and cancellous (trabecular) tissue. Both types are morphologically lamellar bone. Cortical tissue relies on osteons for cell communication. Because trabecular width is small, however, the canaliculi can communicate directly with blood vessels in the medullary canal. The basic differences between cortical tissue and cancellous tissue relate to porosity and apparent density. The porosity of cortical tissue typically ranges from 5% to 30%, and that of cancellous tissue ranges from 30% to 90%. The apparent density of cortical tissue is approximately 1.8 g/cm, and that of cancellous tissue typically ranges from 0.1 to 1.0 g/cm. The distinction between cortical tissue and cancellous tissue is arbitrary, however, and in biomechanical terms, the two tissues are often considered as one material with a specific range in porosity and density.
The organization of cortical and cancellous tissue in bone allows for adaptation to function. Cortical tissue always surrounds cancellous tissue, but the relative quantity of each type of tissue varies with the functional requirements of the bone. In long bones, the cortical tissue of the diaphysis is arranged as a hollow cylinder to best resist bending. The metaphyseal region of the long bones flares to increase the bone volume and surface area in a manner that minimizes the stress of joint contact. The cancellous tissue in this region provides an intricate network that distributes weight-bearing forces and joint reaction forces into the bulk of the bone tissue.
The mechanical properties of cortical bone differ from those of cancellous bone. Cortical bone is stiffer than cancellous bone. Cortical bone fractures in vivo when the strain exceeds 2%, but cancellous bone does not until the strain exceeds 75%. The larger capacity for energy storage (area under the stress–strain curve) of cancellous bone is a function of porosity. Despite different stiffness values for cortical and cancellous bone, the following axiom is valid for all bone tissue: Because the compressive strength of the tissue is proportional to the square of the apparent density, and the elastic modulus or material stiffness of the tissue is proportional to the cube of the apparent density, any increase in porosity, as occurs with aging, will decrease the apparent density of bone, which, in turn, will decrease the compressive strength and elastic modulus of bone.
Variations in the strength and stiffness of bone also result from specimen orientation (longitudinal versus transverse) and loading configuration (tensile, compressive, or shear). Generally, the strength and stiffness of bone are greatest in the direction of the common load application (longitudinally for long bones). With regard to orientation, cortical bone (Figure 1–8) is strongest in the longitudinal direction. With regard to loading configuration, cortical bone is strongest in compression and weakest in shear.
Tensile loading is the application of equal and opposite forces (loads) outward from the surface. Maximal stresses are in a plane perpendicular to the load application and result in elongation of the material. Microscopic studies show that the tensile failure in bones with Haversian systems is caused by debonding of the cement lines and pullout of the osteons. Bones with a large percentage of cancellous tissue demonstrate trabecular fracture with tensile loading.
The converse of tensile loading is compressive loading, which is defined as the application of equal and opposite forces toward the surface. Under compression, a material shortens and widens. Microscopic studies show that compressive failure occurs by oblique cracking of the osteons in cortical bone and of the trabeculae in cancellous bone. Vertebral fractures, especially associated with osteoporosis, are associated with compressive loading.
The application of either a tensile load or a compressive load produces a shear stress in the material. Shear loading is the application of a load parallel to a surface, and the deformation is angular. Clinical studies show that shear fractures are most common to regions with a large percentage of cancellous bone, such as the tibial plateau.
Bone is a viscoelastic material, and its mechanical behavior is therefore influenced by strain rate. Bones are approximately 50% stiffer at high strain rates than at low strain rates, and the load to failure nearly doubles at high strain rates. The result is a doubling of the stored energy at high strain rates. Clinical studies show that the loading rate influences the fracture pattern and the associated soft-tissue damage. Low strain rates, characterized by little stored energy, result in undisplaced fractures and no associated soft-tissue damage. High strain rates, however, are associated with massive damage to the bone and soft tissue owing to the marked increase in stored energy.
Bone fractures can be produced either from a single load that exceeds the ultimate tensile strength of the bone or from repeated loading that leads to fatigue failure. Because bone is self-repairing, fatigue fracture occurs only when the rate of microdamage resulting from repeated loading exceeds the intrinsic repair rate of the bone. Fatigue fractures are most common during strenuous activity when the muscles have become fatigued and are therefore unable to store energy adequately and absorb the stress imposed on the bone. When the muscles are fatigued, the bone is required to carry the increased stress.
Bone has the ability to alter its size, shape, and structure in response to mechanical demands. According to Wolff's law regarding bone remodeling in response to stress, bone resorption occurs with decreased stress, bone hypertrophy occurs with increased stress, and the planes of increased stress follow the principal trabecular orientation. Thus, bone remodeling occurs under a variety of circumstances that alter the normal stress patterns. Clinically, altered stress patterns resulting from fixation devices or joint prostheses have caused concern about effects on the long-term bone architecture.
Bone mass and body weight are positively correlated, especially for weight-bearing bones. Therefore, immobilization or weightlessness (as experienced by astronauts) decreases the strength and stiffness of bone. The subsequent loss in bone mass results from the alteration or absence of normal stress patterns. Bone mass, however, is regained with the return of normal stress patterns. The loss of bone mass in response to immobilization or weightlessness is a direct consequence of Wolff's law. Associated bone resorption in response to orthopedic implants can be deleterious to bone healing, however. Although bone plates provide support for fractured bone, the altered stress patterns associated with stiff metal plates cause resorption of bone adjacent to the fracture or underneath the plate. Therefore, removal of the plate may precipitate another fracture. Resorption of bone is also reported in total hip and knee replacements. This is particularly common with larger diameter noncemented femoral stems, which have an increased moment of inertia and thus have less flexibility than do smaller diameter cemented stems.
The resorption of bone in response to a stiff implant, which alters the stress pattern the bone carries, is termed stress shielding.The degree of stress shielding does not depend on the absolute flexibility of the prosthesis but, rather, on the amount of reduced flexibility in the implant in relation to the flexibility of the bone. Clinically, stress shielding could also be detrimental to the longevity of implant fixation. In an effort to reduce stress shielding, designers of implants are using materials with a degree of flexural rigidity that approximates the flexibility of bone.
The fracture healing process involves five stages: impact, inflammation, soft callus formation, hard callus formation, and remodeling. Impact begins with the initiation of the fracture and continues until energy has completely dissipated. The inflammation stage is characterized by hematoma formation at the fracture site, bone necrosis at the ends of the fragments, and an inflammatory infiltrate. Granulation tissue gradually replaces the hematoma, fibroblasts produce collagen, and osteoclasts begin to remove necrotic bone. The subsidence of pain and swelling marks the initiation of the third, or soft callus, stage. This stage is characterized by increased vascularity and abundant new cartilage formation. The end of the soft callus stage is associated with fibrous or cartilaginous tissue uniting the fragments. During the fourth, or hard callus, stage, the callus converts to woven bone and appears clinically healed. The final stage of the healing process involves slow remodeling from woven to lamellar bone and reconstruction of the medullary canal.
Three types of fracture healing are described. The first type, endochondral fracture healing, is characterized by an initial phase of cartilage formation, followed by the formation of new bone on the calcified cartilage template. The second type, membranous fracture healing, is characterized by bone formation from direct mesenchymal tissue without an intervening cartilaginous stage. Combinations of endochondral healing and membranous healing are typical of normal fracture healing. The former process is observed between fracture gaps, whereas the latter is observed subperiosteally. The third type of fracture healing, primary bone healing, is observed with rigid internal fixation and characterized by the absence of visible callus formation. The fracture site is bridged by direct Haversian remodeling, and there are no discernible histologic stages of inflammation or soft and hard callus formation.
Articular cartilage is primarily avascular and has an abnormally small cellular density. The chief functions of articular cartilage are to distribute joint loads over a large area and to allow relative movement of the joint surfaces with minimal friction and wear.
Articular cartilage is composed of chondrocytes and an organic matrix. The chondrocytes account for less than 10% of the tissue volume, and they manufacture, secrete, and maintain the organic component of the cellular matrix. The organic matrix is a dense network of type II collagen in a concentrated proteoglycan solution. Collagen accounts for 10–30% of the organic matrix; proteoglycan accounts for 3–10%; and water, inorganic salts, and matrix proteins account for the remaining 60–87%.
The basic collagen unit consists of tropocollagen molecules, which form covalent cross-links between collagen molecules to increase the tensile strength of the fibrils. The most important mechanical properties of the collagen fiber are tensile strength and stiffness. Fiber resistance to compression is relatively ineffective because the large ratio of length to diameter (slenderness ratio) predisposes the fibers to buckling. The anisotropic nature of cartilage is thought to be related to several factors, including variations in fiber arrangements within the planes parallel to the articular surface, the collagen fiber cross-link density, and the collagen–proteoglycan interactions.
The mechanical properties of the cartilage are attributed to the inhomogeneous distribution of collagen fibrils (Figure 1–9). The superficial tangential zone contains sheets of fine, densely packed collagen fibers that are randomly woven in planes parallel to the articular surface. The middle zone contains randomly oriented and homogeneously dispersed fibers that are widely spaced to account for increased matrix content. Finally, the deep zone contains larger, radially oriented collagen fiber bundles that eventually cross the tidemark, enter the calcified cartilage, and anchor the tissue to the underlying bone.
Proteoglycans are monomers that consist of a protein core with glycosaminoglycan units (either keratan sulfate or chondroitin sulfate units) covalently bound to the core. Proteoglycan aggregation promotes immobilization of the proteoglycans within the collagen network and adds structural rigidity to the matrix. There are numerous age-related changes in the structure and composition of the proteoglycan matrix, including the following: a decrease in proteoglycan content from approximately 7% at birth to half that by adulthood, an increase in protein content with maturity, a dramatic drop in the ratio of chondroitin sulfate to keratan sulfate with aging, and a decrease in water content as proteoglycan subunits become smaller with aging. The overall effect is that the cartilage stiffens. The development of osteoarthritis is associated with dramatic changes in cartilage metabolism. Initially, there is increased proteoglycan synthesis, and the water content of osteoarthritic cartilage is actually increased.
The water content of normal cartilage permits the diffusion of gases, nutrients, and waste products between the chondrocytes and the nutrient-rich synovial fluid. The water is primarily concentrated (80%) near the articular surface and decreases in a linear fashion with increasing depth, such that the deep zone is 65% water. The location and movement of water are important in controlling mechanical function and lubrication properties of the cartilage.
Important structural interactions occur between proteoglycans and collagen fibers in cartilage. A small percentage of the proteoglycans may serve as a bonding agent between the collagen fibrils that span distances too great for the maintenance or formation of cross-links. These structural interactions are thought to provide strong mechanical interactions. In essence, the proteoglycans and collagen fibers interact to form a porous, composite, fiber-reinforced matrix, possessing all the essential mechanical characteristics of a solid that is swollen with water and able to resist the stresses and strains of joint lubrication.
The biomechanical behavior of articular cartilage is best understood when the cartilage is considered as a viscoelastic and composite material consisting of a fluid phase and a solid phase. The compressive behavior of cartilage is primarily caused by the flow of interstitial fluid, whereas the shear behavior of cartilage is primarily caused by the motion of collagen fibers and proteoglycans. The creep behavior of cartilage is characterized by the exudation of interstitial fluid, which occurs with compressive loading. The applied surface load is balanced by the compressive stress developed within the collagen–proteoglycan matrix and the frictional drag generated by the flow of the interstitial fluid during exudation. Typically, human cartilage takes 4–16 hours to reach creep equilibrium, and the amount of creep is inversely proportional to the square of the tissue thickness.
Similar to creep, stress relaxation is the response of the tissue to compressive forces on the articular surface. An initial compressive phase, characterized by increased stress, is associated with fluid exudation. In the subsequent relaxation phase, stress decay is associated with fluid redistribution within the porous collagen–proteoglycan matrix. The rate of stress relaxation is used to determine the permeability coefficient of the tissue, and the equilibrium stress is used to measure the intrinsic compressive modulus of the solid matrix. Microstructural changes in osteoarthritic cartilage reduce the compressive stiffness of cartilage.
Under uniaxial tension, articular cartilage demonstrates anisotropic and inhomogeneous properties. The tissue is stronger and stiffer parallel to the split lines and in superficial regions. Variations in the material characteristics are a result of the structural organization of the collagen–proteoglycan matrix in layering arrangements throughout the tissue. For example, the superficial tangential zone appears to provide a tough, wear-resistant, protective zone for the tissue. To examine the tissue's intrinsic response to tension, the biphasic viscoelastic effects of the tissue must be negated. This can be achieved by testing the tissue at low strain rates or by performing incremental testing and allowing for stress relaxation equilibrium to be achieved before continuing. The tissue tends to stiffen with increasing strain. Typically, specimens are pulled to the failure point at a displacement rate of 0.5 cm/min.
The shape of the stress–strain curve (Figure 1–10) can be described in morphologic changes of the collagen fibers: (1) the toe region designates collagen fiber pullout, (2) the linear region designates stretching of the aligned collagen fibers, and (3) failure is the point at which all of the collagen fibers have ruptured. The tensile properties of the tissue are thus changed by an alteration of the molecular structure of collagen, an alteration in the organization of the fibers within the collagenous network, or a change in collagen fiber cross-linking. For this reason, disruption of the collagen network may be a key factor in the initial development of osteoarthritis.
When the cartilage is tested in pure shear under infinitesimal strain conditions, no pressure gradients or volume changes are observed within the tissue as they are during tension or compression conditions. Thus, the viscoelastic shear properties of cartilage can be determined in a steady-state dynamic shear experiment. Cartilage shear stiffness is a function of collagen content or collagen–proteoglycan interaction. Increased collagen content reduces frictional dissipation of the load, and this in turn results in increased shear loading.
Sophisticated lubrication processes are responsible for the minimal wear of normal cartilage under large and varied joint stresses. Four types of lubrication mechanisms are related to articular cartilage: boundary, fluid film, mixed, and self-lubrication. These mechanisms are inherent properties of the composition of the tissue with respect to water content and collagen–proteoglycan matrix orientation. Normal joints display all of the lubrication mechanisms just mentioned, whereas artificial joints are thought primarily to display elastohydrodynamic and boundary lubrication mechanisms.
The boundary mechanism protects the joint from surface-to-surface wear by means of an adsorbed lubricant. This mechanism, which depends chiefly on the chemical properties of the lubricant, is most important under severe loading conditions when contact surfaces must sustain high loads.
The fluid film mechanism relies on a thin layer of lubricant that causes greater surface separation. The load on the joint surface is supported by the pressure on the film. Fluid film lubrication occurs with rigid (squeeze-film or hydrodynamic) bodies as well as with deformable (elastohydrodynamic) bodies. When two rigid surfaces are nonparallel and move tangentially with respect to each other, the pressure generated by the lubricant in the gap between the two surfaces is sufficient to raise one surface above the other. Moreover, when two rigid surfaces are parallel and move perpendicular to each other, the pressure generated by the lubricant is sufficient to keep the surfaces separated. This squeeze-film or hydrodynamic lubrication mechanism is able to carry high loads for short durations. When the squeeze-film mechanism generates a pressure great enough to deform the surface and thereby increase the amount of bearing surface area, elastohydrodynamic lubrication mechanisms begin to make the necessary adjustments. Increased bearing surface area allows less lubricant to escape from between the surfaces, decreasing the stress and increasing the duration associated with motion.
The mixed lubrication mechanism is a combination of the boundary and fluid film mechanisms. Boundary lubrication is essential in areas of asperity contact, and fluid film lubrication is present in areas of no contact. Therefore, most of the friction is generated in the boundary lubricated areas, whereas most of the load is carried by the fluid film.
Self-lubrication, or weeping, relies on the exudation of fluid in front of and beneath the surface of the rotating joint. Once the area of peak stress passes a given point, the cartilage reabsorbs the fluid and returns to its original dimensions. This lubrication mechanism results from the inhomogeneous character of the collagen and water distribution throughout the cartilage. When the pressure rises and strains are low, the tissue is most permeable and a large amount of water is exuded in front of the leading contact edge of the joint. As the joint advances, the load increases in the region of expelled fluid and the increased pressure and strains decrease the tissue permeability to fluid. This prevents the fluid on the articular surface from returning to the cartilage. As the contact surface moves past the point of contact, the pressure and strains are again low and the tissue permeability is increased, resulting in the return of fluid to the cartilage in preparation for the cycle to start again.
Synovial fluid is a key component in the lubrication process in joints. It is non-Newtonian, indicating that the resistance to flow increases more slowly than the flow of the fluid. Joint fluid in normal joints, abnormal joints, and artificial joints is different. Hence, different mechanisms of lubrication come into play in each situation with implications for joint surface damage and artificial joint wear.
Wear is the removal of material from a surface and is caused by the mechanical action of two surfaces in contact. The principal types of wear experienced in articular cartilage are interface wear and fatigue wear.
Interface wear occurs when bearing surfaces come into direct contact with no lubricating film separating them. This type of wear may be found in an impaired or degenerated synovial joint. When ultrastructural surface defects in articular cartilage result in softer tissue with increased permeability, the fluid from the lubricant film may easily leak through the cartilage surface, thereby increasing the probability of direct contact between asperities. There are two forms of interface wear: adhesive wear, which occurs when surface fragments adhere to one another and are torn from the surface during sliding, and abrasive wear, which occurs when a soft material is scraped by a harder one.
Fatigue wear results from the accumulation of microscopic damage within the bearing material under repetitive stress. In the cartilage, three mechanisms are primarily responsible for fatigue wear. First, repetitive stress on the collagen–proteoglycan matrix can disrupt the collagen fibers, the proteoglycan molecules, or the interface between the two. In this case, cartilage fatigue is caused by the tensile failure of the collagen network, and proteoglycan changes could be considered part of the accumulated tissue damage. Second, repetitive and massive exudation and inhibition of interstitial fluid may cause a proteoglycan washout from the cartilage matrix near the articular surface. This results in decreased stiffness and increased tissue permeability. Third, during synovial joint impact loading, insufficient time for internal fluid redistribution to relieve high stress in the compacted region may result in tissue damage.
Numerous structural defects of the articular cartilage are caused or exacerbated by wear and damage. For example, fibrillations (splitting of the articular surface) are associated with wear and will eventually extend the full thickness of the cartilage. Destructive smooth-surface thinning is apparent when layers erode rather than split. In these and other types of surface damage of the cartilage, more than a single wear mechanism is likely to be responsible.
Several biomechanical hypotheses cover cartilage degradation. Factors associated with progressive failure of the tissue include the magnitude of imposed stress, the total number of sustained stress peaks, changes in the intrinsic molecular and microscopic structure of the collagen–proteoglycan matrix, and changes in the intrinsic mechanical property of the tissue. Failure-initiating mechanisms include a loosening of the collagen network, which allows for abnormal expansion of the proteoglycan matrix and swelling of the tissue, and a decrease in cartilage stiffness, which is accompanied by an increase in tissue permeability.
Biomechanically, conditions that cause excessive stress concentrations may result in increased tissue damage or wear. Joint surface incongruity, such as the incongruity of the hip joint in patients who had Perthes disease during childhood, can result in abnormally small contact areas, which are associated with increased stress and increased tissue damage. Moreover, the presence of high contact pressures between the articular surfaces, such as that seen in patients with a shallow acetabulum (acetabular dysplasia), can reduce the probability of fluid film lubrication, allow for continued tissue damage, and also increase the risk of early degenerative arthritis.
Tendons & Ligaments
Tendons and ligaments are similar both structurally and biomechanically and differ only in function. Tendons attach muscle to bone; transmit loads from the muscle to the bone, which results in joint motion; and allow the muscle belly to remain an optimal distance from the joint on which it acts. Ligaments attach bone to bone, augment mechanical stability of the joint, guide joint motion, and prevent excessive joint displacement.
Both the tendons and the ligaments are parallel-fibered collagenous tissues that are sparsely vascularized. They contain relatively few fibroblasts (constituting approximately 20% of their volume) and an abundant extracellular matrix. The matrix consists of approximately 70% water and 30% collagen, ground substance, and elastin.
The fibroblasts secrete a precursor of collagen, procollagen, which is cleaved extracellularly to form type I collagen. Cross-links between collagen molecules provide strength to the tissue. The arrangement of the collagen fibers determines tissue function. In tendons, a parallel arrangement of the collagen fibers provides the tissues with the ability to sustain high uniaxial tensile loads. In ligaments, the nearly parallel fibers, which are intimately interlaced with one another, provide the ability to sustain loads in one predominant direction but allow for carrying small tensile loads in other directions.
Tendons and ligaments are surrounded by loose areolar connective tissue. The paratenon forms a protective sheath around the tissue and enhances gliding. At places where the tendons are subjected to large friction forces, a parietal synovial membrane is found just beneath the paratenon and additionally facilitates gliding. Each individual fiber bundle is bound by the endotenon. At the musculotendinous junction, the endotenon continues into the perimysium. At the tendoosseous junction, the collagen fibers of the endotenon continue into the bone as perforating fibers (Sharpey fibers) and become continuous with the periosteum.
Tendons and connective tissues of the musculotendinous junction help determine the mechanical characteristics of whole muscle during contraction and passive extension. The muscle cells are extensively involuted and folded at the junction to provide maximal surface area for attachment, thereby allowing for greater fixation and transmission of forces. The sarcomeres directly adjacent to the junction of fast contracting muscles are shortened in length. This may represent an adaptation to decrease the force intensity within the junction. A complex intracellular and extracellular transmitting membrane consisting of a glycoprotein links the contractile intracellular proteins to the extracellular protein connective tissue.
The tendon insertions and ligament insertions to the bone are structurally similar. The collagen fibers from the tissue intermesh with fibrocartilage. The fibrocartilage gradually becomes mineralized, and this mineralized cartilage merges with cortical bone. These transition zones produce a gradual alteration in the mechanical properties of the tissue, resulting in a decreased stress concentration effect at the insertion of the tendon or ligament to the bone.
Tendons and ligaments are viscoelastic structures that have specific mechanical properties related to their function and composition. Tendons are strong enough to sustain high tensile forces resulting from muscle contraction during joint motion, but they are also sufficiently flexible to angulate around bone surfaces, to change the final direction of muscle pull. Ligaments are pliant and flexible enough to allow natural movements of the bones they connect; however, they are strong, are not extensible, and offer suitable resistance to applied forces and large joint movements. Because tendons and ligaments are viscoelastic structures, the injury they sustain is affected by the rate of loading as well as the amount of the stress load. The stress–strain and load-elongation curves for ligaments and tendons, like those for articular cartilage, have several regions that characterize the tissue behavior.
Figure 1–11 shows the load-elongation curve for progressive failure of the anterior cruciate ligament. Like the curve in Figure 1–10, the curve in Figure 1–11 has a toe region (correlating with the region labeled clinical test, when the anterior drawer test was administered) and a linear region preceding the failure region. In Figure 1–11, the curve in the toe region represents large elongations with small changes in load. This pattern is thought to reflect the straightening of the wavy, relaxed collagen fibers with increased loads. Within the linear region, the collagen fibers continue to become more parallel in orientation as physiologic loading proceeds. At the end of the linear region, small force reductions can be observed in the load-deformation curve. These dips are caused by the early sequential failure of a few maximally stretched fiber bundles. The final region represents major failure of fiber bundles in an unpredictable manner. Complete failure occurs rapidly, and the load-supporting ability of the tissue is substantially reduced.
The mechanical behavior characteristics of the anterior cruciate ligament differ somewhat from those of soft tissues that contain a high proportion of elastin fibers. These tissues can elongate up to 50% before stiffness markedly increases. After 50% elongation, however, the stiffness increases greatly with increased loading, and failure is abrupt with minimal further elongation. Load-elongation curves for several soft tissues are shown in Figure 1–12.
The viscoelastic behavior of ligaments is best exemplified in the bone-ligament-bone complex. Anterior cruciate ligaments in primate knee specimens were tested in tension to failure at both slow and fast loading rates to determine the viscoelastic nature of the bone-ligament-bone complex. At slow loading rates the bony insertion of the ligament was the weakest link, and an avulsion resulted. At fast loading rates, the ligament was the weakest link, and a midsubstance rupture generally was found. At slow rates, the load to failure was decreased by 20% and the stored energy was decreased by 30% in comparison with results with fast rates. The stiffness of the bone-ligament-bone complex was relatively unaffected by strain rate, however. Increased strain rates demonstrated a greater increase in strength for bone as compared with ligaments.
The mechanical properties of ligaments are closely related to the number and quality of the cross-links within the collagen fibers. Therefore, any process that affects collagen formation or maturation directly influences the properties of the ligaments. As aging continues, the number and the quality of cross-links increase, thereby increasing the tensile strength of the tissue. Moreover, the diameter of the collagen fibril increases with age. As aging progresses, however, collagen reaches a mechanical plateau, after which point tensile strength and stiffness decrease. There is also a decrease in the tissue collagen content, and this contributes to the continued decline in the mechanical properties of the tissue.
Tendons and ligaments remodel in response to mechanical demand. Physical training increases the tensile strength of the tendons and the ligament-bone interface, whereas immobilization decreases tensile strength. Even if the tissue maintains a relatively constant cross-sectional area during immobilization, the increased tissue metabolism results in proportionately more immature collagen and a decrease in the amount and quality of cross-links between molecules. Investigators who studied ligaments that were immobilized for 8 weeks and control ligaments found that the previously immobilized ligaments required 12 months of reconditioning before they demonstrated strength and stiffness values comparable to those of the control ligaments.
Studies of nonsteroidal antiinflammatory drugs (NSAIDs) such as indomethacin demonstrated that treatment results in increases in the proportion of insoluble collagen and the total collagen content in tissue. It also leads to increased tensile strength, which is probably attributable to increased collagen molecule cross-links. Therefore, short-term NSAID therapy may increase the rate of biomechanical restoration of the tendons and ligaments.
Tendons and ligaments are subjected to less than a third of their ultimate stress during normal physiologic loading. The maximal physiologic strain ranges from 2% to 5%. Several factors lead to tissue injury, however. When tendons and ligaments are subjected to stresses that exceed the physiologic range, microfailure of collagen bundles occurs before the yield point of the tissue is reached. When the yield point is reached, the tissue undergoes gross failure and the joint simultaneously becomes displaced. The amount of force produced by the maximal contraction of the muscle results in a maximal tensile stress in the tendon. The extent of tendon injury is influenced by the amount of tendon cross-sectional area compared with that for muscle. The larger the muscle cross-sectional area, the higher the magnitude of the force produced by the contraction and thus the greater the tensile load transmitted through the tendon.
Clinically, ligament injuries are characterized according to degree of severity. First-degree sprains are typified by minimal pain and demonstrate no detectable joint instability despite microfailure of collagen fibers. Second-degree sprains cause severe pain and demonstrate minimal joint instability. This instability is most likely masked by muscle activity, however. Therefore, testing must be performed with the patient under anesthesia for proper evaluation. Second-degree sprains are characterized by partial ligament rupture and progressive failure of the collagen fibers, with the result that ligament strength and stiffness decrease by 50%. Third-degree sprains cause severe pain during the course of the injury and minimal pain afterward. The joint is completely unstable. Most collagen fibers have ruptured, but a few may remain intact, giving the ligament the appearance of continuity even though it is incapable of supporting loads. Abnormally high stress on the articular cartilage results if pressure is exerted on a joint that is unstable owing to ligament or joint capsule rupture.
During tendon and ligament healing and repair, fibroblastic infiltration from the adjacent tissues is essential. The healing events are initiated by an inflammatory response, which is characterized by polymorphonuclear cell infiltration, capillary budding, and fluid exudation and continues during the first 3 days following the injury. After 4 days, fibroplasia occurs and is accompanied by the significant accumulation of fibroblasts. Within 3 weeks, a mass of granulation tissue surrounds the damaged tissue. During the next week, collagen fibers become longitudinally oriented. During the next 3 months, the individual collagen fibers form bundles identical to the original bundles.
Sutured tendons heal with a progressive penetration of connective tissue from the outside. The deposited collagen fibers become progressively oriented until eventually they form tendon fibers like the original ones. This orientation of collagen fibers is essential because the tensile strength of repaired tendon depends on collagen content and orientation. If tendon is sutured during the first 7–10 days of healing, the strength of the suture maintains the fixation until adequate callus forms.
Tendon mobilization during healing is important to avoid adhesion of the tendon to adjacent tissue, particularly in cases involving the flexor tendons of the hand. Motion can be passive to prevent adhesion and at the same time to prevent putting excessive tensile stress on the suture line. The gliding properties of flexor tendons that were mobilized are consistently superior to those of flexor tendons that were immobilized during the healing process.
Direct apposition of the surfaces of a divided ligament provides the most favorable conditions for healing because it minimizes scar formation, accelerates repair, hastens collagenization, and comes closer to restoring normal ligamentous tissue. Care must be taken during the repair of ligaments to avoid subsequent common problems with healing, however. For instance, divided and immobilized ligaments heal with a fibrous tissue gap between the two ends, whereas sutured ligaments unite without a fibrous tissue gap. If excessive tension is placed on a suture, necrosis and failure to heal are observed. Unsutured ligaments can retract, shorten, and become atrophic, however, making repair difficult 2 weeks following the injury. In spite of this, many ligaments are not routinely repaired in orthopedic surgery.
The anterior cruciate ligament is often severely damaged in cases of midsubstance rupture and generally does not fare well following repair. The ligament is intraarticular, with synovial fluid tending to disrupt the repair. Instability of the knee also tends to place excessive stress on the repair unless the knee is immobilized, which leads to joint stiffness and muscle atrophy.
Skeletal muscles perform a wide variety of mechanical and biologic functions. From a mechanical perspective, it is obvious that skeletal muscles generate force and length changes. The generation of force and length change gives rise to the production of mechanical work and power. Less obvious is the fact that skeletal muscles are often subjected to so-called lengthening or eccentric contractions. During these types of contractions, muscles may act as so-called dynamic joint stabilizers and may store energy. From a biologic perspective, skeletal muscles are believed to secrete various growth factors such as insulin-like growth factor 1 (IGF-1), which is thought to play an important autocrine/paracrine role in regulating muscle fiber size. Additionally, it has been proposed that skeletal muscles play a key role in maintaining the health of motor neurons.
SKELETAL MUSCLE STRUCTURE
Figure 1–13 provides both a macroscopic and microscopic perspective of the structure of skeletal muscle. From a macroscopic perspective, skeletal muscles are composed of tens of thousands of individual muscle fibers (muscle cells). Muscles that are involved in fine motor control usually contain a small number of muscle fibers compared with those muscles involved in activities requiring the generation of large forces and power outputs. Muscle fibers are usually found in so-called bundles that are also referred to as fascicles. Each fascicle typically contains approximately 10–30 muscle fibers that are encased in a connective tissue sheath known as the endomysium.
From an architectural perspective, muscles are often classified on the basis of the orientations of the muscle fibers' longitudinal axes relative to that of the entire muscle. For instance, longitudinal muscles are composed of muscle fibers whose longitudinal axis runs parallel to that of the whole muscle. Good examples of this type of architecture are the rectus abdominis and the sartorius muscles. In fusiform muscles, the fibers run parallel to the longitudinal axis throughout most of the muscle, but they taper at the ends of the muscle. The soleus and brachioradialis muscles are typical of this architecture. Muscles can also exhibit a so-called pennate(unipennate, bipennate) architecture whereby the longitudinal axis of the individual muscle fibers runs diagonal to that of the whole muscle. A good example of a bipennate muscle is the gastrocnemius muscle. The muscle fibers of angular or fan-shaped muscles radiate from a narrow attachment at one end and fan out, resulting in a broad attachment at the other end as is seen in muscles like the pectoralis major.
Consistent with the theme of structure–function relationships, muscle architecture can be an important determinant of the mechanical properties of skeletal muscle. For instance, fusiform muscles typically have longer muscle fibers than bipennate muscles. Functionally, this means that a fusiform muscle should be able to generate greater shortening velocities and muscle length excursions at the whole muscle level. In contrast, muscles with a pennate or bipennate architecture have shorter fibers, but the fibers are packed in such a manner that a larger number of muscle fibers are in parallel to one another, resulting in a larger physiologic cross-sectional area. Hence, the pennate muscle has a greater capacity for generating force.
Molecular Anatomy of the Myofibril
The structure of skeletal muscle at the molecular level is quite complex (see Figure 1–13). Each muscle fiber is made up of thousands of so-called myofibrils that are arranged in parallel to one another. Each myofibril has a cross-sectional area of approximately 1 m2. Hence, a muscle fiber with a cross-sectional area of approximately 1000 m2 would contain approximately 1000 myofibrils. Typically, the cross-sectional area of a muscle fiber can range from approximately 1000 to 7000 m2. Each myofibril consists of a repeating series of striations that are caused by the arrangement of so-called sarcomeres in series. Each sarcomere is approximately 2–3 m in length. Sarcomeres are often referred to as the contractile units of skeletal muscle.
In a general sense, sarcomeres consist of Z-lines, thin filaments, and thick filaments. The interdigitation of thick and thin filaments along with the presence of Z-lines is primarily responsible for the striation pattern of skeletal muscle. As shown in Figure 1–14, the Z-lines are dense thin structures that are found in the middle of the so-called I-band. In reality, each Z-line represents an anchor point to which thin filaments are attached. By definition, the collection of proteins between each Z-line is known as a sarcomere. Hence, the I-band represents a region where no overlap occurs of the thin filaments (by thick filaments), yielding a relatively light band. The A-band is composed of the thick filament and is strongly birefringent, producing a dark band on microscopic inspection. By definition, the length of the A-band is equivalent to the length of the thick filament. Normally, the thick and thin filaments partially overlap, and as a result a lighter region occurs in the middle of the A-band known as the H-zone.
Changes in sarcomere length and, as a result, muscle fiber length are caused by the sliding of the thick and thin filaments relative to one another. In its most simplistic sense, this model states that contraction takes place not because of changes in the individual lengths of thick and thin filaments, but rather by the sliding of thin filaments past thick filaments. This model of contraction is known as the sliding-filament hypothesis. The changes in striation patterns during shortening contractions played a central role in developing the sliding-filament hypothesis. In this context, Table 1–1 summarizes the changes in the striation pattern that occur during isometric, shortening (isotonic), and lengthening (eccentric) contractions.
Molecular Anatomy of the Sarcomere
As shown in Figure 1–14 and in Table 1–2, the overall structure of the sarcomere became quite complex as more sophisticated techniques for studying skeletal muscle evolved. On a basic level, the sarcomeric proteins and those associated with the sarcomere can be placed into four different categories: (1) contractile; (2) regulatory contractile; (3) structural; and (4) costameric.
As shown in Table 1–2, the primary contractile proteins are simply actin and myosin. These are referred to as contractile proteins, given their central role in the contractile process. Individual monomers of actin bind to one another to form so-called actin filaments. In contrast, the thick filament is composed primarily of myosin heavy-chain molecules packed in an antiparallel arrangement. A more detailed description of myosin is provided later. Regulatory contractile proteins are defined as those that turn the contractile apparatus on or off and those that can modulate the activity of the myosin heavy chain. In skeletal muscle, the regulatory contractile proteins involved in turning the contractile apparatus on or off are associated exclusively with the actin filament, and these proteins include tropomyosin, troponin-T, troponin-I, and troponin-C. Collectively, the thin filament is composed of the actin filament and these (ie, tropomyosin, troponin-T, troponin-I, and troponin-C) regulatory contractile proteins. Other regulatory contractile proteins are associated with the myosin heavy chain, and these are referred to collectively as myosin light chains (MLCs) because of their relatively low molecular weight. These MLCs may possibly be involved in regulating the kinetics of crossbridge cycling.
Structural and costameric proteins play several essential roles. First, electron micrographs (see Figure 1–14) demonstrate that sarcomeres are organized in a very orderly fashion such that the Z-lines of adjoining sarcomeres appear to be in register with one another. As noted in Table 1–2, key intermediate filaments like desmin and vimentin are believed to play key roles in aligning the Z-line of one sarcomere with that of another. Other proteins like synemin are also thought to be involved in the alignment of sarcomeres. Structural proteins also play a key role in developing a mechanical linkage between sarcomeres and the extracellular matrix. These sites of connectivity between the sarcomere, cell membrane, and extracellular matrix are referred to as costameres(see Molecular Anatomy of the Connection between the Cytoskeleton and the Extracellular Matrix).
Molecular Anatomy of Myosin Molecule
Although the term myosin molecule is often used, in reality the myosin molecule is a hexameric structure composed of two myosin heavy chains, two so-called essential light chains (MLC1 or MLC3), and two regulatory light chains (MLC2). The term heavy or light is used in reference to the molecular weights of each of these proteins. Each myosin heavy chain is composed of a rod region, lever arm (also known as S2), and a globular head (also known as S1). The rod region plays an important role in the packing of individual myosin heavy chains into thick filaments. The globular head contains the key functional domains of this molecular motor. Within the globular head (Figure 1–15) are domains that contain (1) the actin-binding site, (2) the nucleotide (adenosine triphosphate [ATP])-binding site, and (3) the enzymatic (adenosine triphosphatase [ATPase]) properties responsible for converting chemical energy in the form of ATP into mechanical work and heat. The essential and regulatory light chains are bound to the so-called lever arm or S2region (see Figure 1–15). Each globular head has bound to it one essential and one regulatory light chain. Mutations of some of these domains (eg, the actin-binding site) are thought to play roles in diseases such as familial hypertrophic cardiomyopathy. As noted earlier, the light chains are thought to play modulatory roles in regulating the kinetics of the crossbridge cycle.
The complexity of the myosin molecule is further complicated by the presence of isoforms for both the myosin heavy and light chains. Although it has long been recognized that muscles could be classified as slow or fast twitch (based on twitch properties), the importance of myosin heavy-chain isoforms has only been intensely studied during the past 20 years. In many smaller adult mammals (eg, mice, rats, rabbits), four myosin heavy-chains isoforms are identified and classified as slow Type I, fast Type IIA, fast Type IIX, and fast Type IIB (in order of increasing ATPase activities and associated maximal shortening velocities). In adult humans, the slow Type I, fast Type IIA, and fast Type IIX myosin heavy-chain isoforms are expressed. This scheme of classifying myosin heavy-chain isoforms forms the basis for the nomenclature typically used to identify different muscle fiber types. Hence, a slow Type I muscle fiber would exclusively express the slow Type I myosin heavy-chain isoform.
Note that there are isoforms for myosin heavy chains, myosin light chains, tropomyosin, troponin-T, troponin-I, and troponin-C. Hence, by mixing and matching these contractile and regulatory contractile proteins, the complexity that can arise in the design of sarcomeres becomes readily apparent.
The sequence of the crossbridge cycle is shown in Figure 1–15. First, myosin is detached from actin. Second, the head of the myosin heavy chain is attached to actin and releases Pi, leading to the power stroke (change in position of head between Figure 1–15B and 1–15C). Following completion of the power stroke, adenosine diphosphate (ADP) is released, and subsequently ATP binds to the nucleotide-binding site (Figure 1–15D, E). The hydrolysis of ATP ultimately leads to the globular head of the myosin heavy chain returning to its original position. The magnitude of crossbridge cycling that occurs during a single contraction is enormous and can approach rates equivalent to 1017–1018 crossbridge cycles per gram of muscle per second.
In thinking about the plasticity of the sarcomere and its constituent proteins, note that mechanical unloading and denervation lead to a decrease in the number of sarcomeres in parallel. From a functional perspective, this leads to a decrease in the capacity to produce force. In contrast, resistance training leads to an increase in the number of sarcomeres in parallel, and, as a consequence, increased capacity to produce force.
Factors such as mechanical unloading (eg, as accompanies cast immobilization) and altered thyroid hormone status produce shifts in the contractile and regulatory contractile protein isoform profiles such that they become faster. For instance, cast immobilization may lead to a transition from the slow type I to the fast type IIX myosin heavy-chain isoform. From a functional perspective, this will lead to an increase in maximal shortening velocity. Although it is commonly thought that strength training is an effective tool for increasing sprint speed, note that this will produce fast-to-slow transitions in myosin heavy-chain isoform expression.
Molecular Anatomy of the Connection between the Cytoskeleton and the Extracellular Matrix
Costameres are important structures that link the cytoskeleton of skeletal muscle with the extracellular matrix (Figure 1–16). Currently, it is believed that the costameres serve at least three different functions: (1) aligning the sarcolemma with the cytoskeleton; (2) maintaining membrane integrity during different types of contractions, preventing injury to the sarcolemma; and (3) possibly playing a role in the lateral transmission of force. From a structural perspective, costameres are found aligned with the Z- and M-lines. This linkage occurs as a result of so-called intermediate filaments. Note that mutations in the costameric structure are thought to be involved in some of the muscular dystrophies. Additionally, these structures may play an important role in protecting muscle fibers from eccentrically induced muscle damage.
The On-and-Off Switch of the Molecular Motor: Excitation-Contraction Coupling
From a mechanical perspective, all of the sarcomeres must become activated in a synchronous fashion. An asynchronous activation of sarcomeres would lead to large heterogeneities in sarcomere length along the length of a muscle fiber with some sarcomeres actively shortening, whereas the nonactivated sarcomeres would be lengthened. The net result might be a contraction, whereby there is little overall shortening of the muscle fiber. The synchronous activation of all sarcomeres requires an elaborate reticulum that functionally couples the depolarization of the sarcolemma (cell membrane) with activation of the sarcomere. The coupling between excitation and contraction involves extensive invaginations of the sarcolemma known as transverse tubules (T-tubules), which are associated with the sarcoplasmic reticulum. The sarcoplasmic reticulum is a network of membranes wrapped around the myofibrils, containing a large store of Ca2+. When a skeletal muscle fiber is excited, the depolarization of the sarcolemma is propagated into the T-tubules. Excitation of the T-tubules then leads to the release of Ca2+ from the ends of the sarcoplasmic reticulum via so-called Ca2+ release channels (also known as ryanodine receptors). The Ca2+ quickly diffuses into the space occupied by the sarcomere, binding to troponin-C. This then causes tropomyosin to rotate about the longitudinal axis of the actin filament, uncovering the myosin binding sites of each actin molecule. The globular head of the myosin heavy chain attaches to this binding site and goes through its power stroke, leading to the production of force or length change. Simply stated it is the binding of Ca2+ to troponin-C that turns on the contractile apparatus.
The contractile activity of the sarcomere is turned off by resequestering Ca2+ back into the sarcoplasmic reticulum via Ca2+ ATPase pumps located along the length of the sarcoplasmic reticulum. The dissociation of Ca2+ from troponin-C causes tropomyosin to rotate back into its original position, once again covering up or blocking the myosin-binding sites of each actin molecule. In this manner, the globular head of the myosin heavy chain is prevented from binding to actin, leading to decay in force production and causing the muscle to relax.
As mentioned earlier, there are a number of isoforms for the contractile and regulatory contractile proteins. Isoforms are also identified for the Ca2+ release channels and Ca2+ ATPase pumps of the sarcoplasmic reticulum. These different isoforms play a key role in determining the rate of activation (ie, the release of Ca2+) and relaxation (resequestration of Ca2+). Mechanical unloading of skeletal muscle typically leads to an increased expression of the fast isoforms of the Ca2+ release channels, whereas reloading or strength training produces the opposite effect.
SKELETAL MUSCLE FUNCTION
In a general sense, the mechanical activity of skeletal muscle depends on two factors: the pattern of stimulation and the extent of loading. The most basic unit of contractile response is known as a twitch. Simply stated, a twitch is the mechanical response of skeletal muscle to a single brief stimulus (Figure 1–17). This single stimulus leads to a single pulse of Ca2+ released from the sarcoplasmic reticulum. This single pulse of Ca2+ is nonsaturating, meaning that it binds to only some of the troponin-C molecules. In turn, this causes only some of the tropomyosin molecules to rotate about the longitudinal axis of the actin filament, uncovering only some of the myosin-binding sites. From a mechanical perspective, this leads to a submaximal force transient called a twitch (see Figure 1–17A). The Ca2+ is quickly resequestered by the sarcoplasmic reticulum, and force returns to resting levels.
The amount of Ca2+ released by the sarcoplasmic reticulum can be modulated by the pattern of stimulation. If a muscle is repetitively stimulated but with long durations between each stimulus, then the mechanical response will appear as a series of individual twitches, and the force produced will be submaximal. If the muscle is stimulated using a moderate frequency, however, then the mechanical response caused by one stimulus will fuse with that of the second stimulus, leading to a mechanical response known as tetanus, and the amplitude depends on the balance between Ca2+ release and resequestration. A greater frequency of stimulation leads to a greater release of Ca2+ and production of force (Figure 1–17B). Tetanus has two types: one in which partial relaxation occurs between each stimulus (unfused) and another in which no discernible relaxation happens between stimuli (fused) (Figure 1–17B). During fused tetanus, all of the Ca2+-binding sites of troponin-C are saturated, causing all of the myosin-binding sites to be exposed. This results in the greatest production of force.
As noted earlier, the loading of skeletal muscle also plays a key role in determining the mechanical response. For instance, if a muscle contracts against an immovable object, the muscle does not shorten, and hence muscle length remains constant. This type of contraction is referred to as an isometric contraction (iso = same; metros = length), and it is under these loading conditions that the muscle produces maximal force. If a muscle contracts against a load that is submaximal (ie, less than the maximal force the muscle can generate), then the muscle shortens. This type of contraction is often referred to as either a shortening or an isotonic contraction. Muscles not only work under these types of loading conditions, but they work almost as often under conditions in which the muscle is activated and forcibly lengthened. These types of contractions are known as lengthening or eccentric contractions.
Conceptual Framework of Factors That Determine Muscle Function
From an orthopedics perspective, it would be beneficial to have a framework that would include the key factors that determine the mechanical function of skeletal muscle. Such a framework is laid out in Figure 1–18 and revolves around the context of net mechanical work. In a general context, the net mechanical work produced by skeletal muscle during cyclic length changes (eg, cyclic sinusoidal length changes) is determined by those factors that determine the positive amount of work produced during the shortening phase and those factors that determine the amount of mechanical work done on relengthening the muscle (ie, negative work).
The four determinants of the amount of positive work that can be produced during the shortening phase are (1) the length–tension relationship, (2) the rate of activation, (3) the force–velocity relationship in the shortening domain, and (4) the rate of relaxation. Two factors determine the amount of work done on the muscle during lengthening: (1) the force–velocity relationship in the lengthening domain and (2) the passive stiffness of the muscle. From an engineering perspective, each of these factors can be thought of as representing design constraints, one of which is static (the length–tension relationship) and others are dynamic. The term static in reference to the length–tension relationship implies that the basic dimensions of the sarcomere (and hence length–tension relationship) do not appear to be malleable. In contrast, the other factors can all be altered by factors influencing the mechanical loading of skeletal muscle (eg, immobilization), innervation (partial/complete denervation), and hormonal milieu (eg, thyroid, steroids).
The amount of force a muscle can generate depends on muscle length; this length–tension relationship is shown in Figure 1–19. Typically, three regions of the length–tension relationship are described. The ascending limb extends from a sarcomere length of approximately 1.3–2.0 m. In this region, the amount of isometric tension increases in direct proportion to the increase in sarcomere length. The plateau region extends from approximately 2.0 to 2.5 m in mammalian fibers, and in this range there is an optimal overlap between the thick and thin filaments. Beyond a sarcomere length of 2.5 m, the isometric force that can be produced decreases as a linear function of increases in sarcomere length, reflecting a progressive decrease in overlap between the thick and thin filaments.
The length–tension relationship of the sarcomere is thought to represent a static design criteria, implying that it does not change with various types of interventions such as mechanical unloading. However, it is well known that muscles immobilized in a lengthened position increase the number of sarcomeres in series, leading to the longitudinal growth of the fiber. In contrast, immobilization in a shortened position reduces the number of sarcomeres in series and results in a shorter muscle fiber. Hence, such manipulations have the potential for influencing the overall length–tension relationships of muscle fibers and whole muscles. Clearly, contractures may have a large effect on the number of sarcomeres in series and result in various clinical complications such as equinus contractures.
Force–Velocity Relationship in the Shortening Domain
When a muscle contracts against a light load, it is able to shorten at a relatively high velocity. However, when a muscle contracts against a heavy load, it shortens at a relatively slow velocity. The relationship between the force and shortening velocity is shown in Figure 1–17C and reveals that the force–velocity relationship in the shortening domain can be described by a rectangular hyperbola. Importantly, the shape and dimensions of this relationship depend on the types of contractile protein isoforms and how they are packaged. As noted earlier, maximal shortening velocity is primarily determined by the types of myosin heavy-chain isoforms present in the muscle fiber. For instance, a muscle fiber that expresses only the slow type I myosin heavy chain has a much slower Vmax than one that expresses only the fast type IIX or IIB myosin heavy-chain isoforms. Vmax does not depend on the cross-sectional area of the muscle fiber. The maximal isometric tension that a muscle can produce, often referred to as Po, is largely independent of myosin heavy-chain isoforms but is heavily dependent on the number of sarcomeres in parallel (ie, cross-sectional area). Hence, in an individual who has marked atrophy, Po would be expected to be significantly reduced but Vmax would be unchanged. In this context, the force–velocity relationship can be referred to as a dynamic design criterion, meaning that changes in either cross-sectional area or myosin heavy-chain isoform expression can alter the shape and dimensions of this relationship. Importantly, this relationship represents a design criterion because the muscle can only operate on or below the force–velocity relationship.
The product of force x velocity is mechanical power. Hence, the force–velocity relationship also defines the maximal amount of mechanical power that can be produced under any given loading condition. This has important implications for a wide variety of movements.
Force–Velocity Relationship in the Lengthening Domain
When a muscle is maximally activated and then forcibly lengthened, the tension that can be generated is much greater than that observed under isometric conditions. However, at velocities beyond a relatively low lengthening velocity, tension does not rise any further. This is shown in Figure 1–20. Although muscles commonly perform lengthening contractions, our understanding of the factors that determine the shape of the force–velocity relationship in the lengthening domain is relatively poor. The shape and dimensions of the force–velocity relationship in the lengthening domain are much more complicated than these same dimensions in the shortening domain. In the shortening domain, the force–velocity relationship can be described by a planar curve with the axes of force and velocity. In the lengthening domain, the relationship is three-dimensional, dependent on force, velocity, and time.
An example of the response of skeletal muscle to a lengthening contraction is shown in Figure 1–17D. Note that the muscle is stimulated, and force is allowed to move onto its isometric plateau. A constant-velocity stretch is then imposed on the muscle, and tension rises rapidly. However, at a strain of approximately 1–2%, a sudden change occurs in the rise in tension such that the slope decreases dramatically. The initial force response is often referred to as short-range stiffness. Note that both before and after the yield, force is constantly changing but the velocity of lengthening remains constant. This demonstrates the complexity of understanding the force–velocity relationship in the lengthening domain.
From a functional perspective, the force–velocity relationship in the lengthening domain is important because it determines the amount of work done on relengthening a muscle that is either fully or partially activated. Additionally, muscles are thought to act as dynamic joint stabilizers, so it might be hypothesized that stiffer muscles might better protect joints and ligaments (ie, anterior cruciate ligament) that are susceptible to injury. Newer studies show that mechanical unloading dramatically reduces both the stiffness and elastic modulus of skeletal muscle. The loss of elastic modulus occurs either because of a decrease in crossbridge density or to changes in the unitary stiffness of crossbridges.
Passive Stiffness of Skeletal Muscle
The passive stiffness of skeletal muscle is influenced by both sarcomeric and extracellular matrix proteins. Titin is the largest known protein identified to date, and it attaches the thick filaments to the Z-line. Within the titin molecule is a unique region, the PEVK region, that functions like a molecular spring. The term PEVK refers to the abundance of proline (P), glutamate (E), valine (V), and leucine (K) found in this region. The passive tension of skeletal muscle fibers during stretch may, in part, be caused by the properties of titin. Newer evidence suggests that titin also acts as a molecular blueprint for the organization and structure of the sarcomere, and that titin may play a key role in regulating mechanosensitivity and mechanotransduction in skeletal muscle. For instance, titin contains a kinase domain near the M-line, and this appears to play a prime role in regulating the response of skeletal muscle to increased/decreased mechanical loading by providing a signaling bridge between the sarcomere and nucleus. Newer data suggest that mechanical unloading of skeletal muscle (especially slow type I fibers) results in a loss of titin relative to other myofibrillar proteins such as myosin, but it does not produce shifts in the types of titin isoforms. This, in turn, is believed to impact the passive stiffness of skeletal muscle, but it should be emphasized that the loss in stiffness associated with mechanical unloading is relatively small.
Although the extracellular matrix (ECM) is composed of a number of proteins, the major constituents are collagen and laminin. Fifteen different collagens are identified to date, and the most prevalent types found in skeletal muscle are types I and III. Type I collagen is more common in slow skeletal muscle, whereas type III is more abundant in fast skeletal muscles. Type I and III collagens are each composed of three individual chains that are arranged in a triple helical fashion. Type I collagen is composed of two 1 chains and one 2 chain. In contrast, type III is composed of three 1 chains. Each of the different types of chains (eg, I1) is a product of its own specific gene. From a mechanical perspective, type I collagen is known to have a high tensile strength, whereas type III is more compliant.
Skeletal muscle fibers can be exposed to a variety of forces (eg, normal and shear stress) and strains (normal and shear strain). In a simplistic sense, isometric contractions are associated with high stresses and no strain. Alternatively, passive stresses can be imposed on a skeletal muscle fiber that result in stretch, producing a given amount of strain (relative change in length). To date, the scientific community only has a basic understanding of the effects of different loading conditions on the collagen content of skeletal muscle and associated regulatory pathways. Several studies show that mechanical unloading via hindlimb suspension results in a down-regulation of the expression of type I and III collagens. In contrast, the responses of collagen genes/proteins to chronic stretch have not been studied in any detail, yet knowledge in this area may have important clinical ramifications. For instance, orthopedic surgeons commonly use distraction osteogenesis (Ilizarov procedure) as a method to correct limb length deformities. Unfortunately, such procedures (ie, distraction osteogenesis) often result in complications like equinus contractures, which are hypothesized to occur in skeletal muscle because of a proliferation of collagen content.
The intervertebral disks sustain and distribute loads and also prevent excessive motion of the spine. An individual's intervertebral disks account for 20–33% of his or her spinal column height. The disks are subjected to high stresses during normal daily activity, and stress may double during increased activity, lifting, or trauma. Whether intervertebral disk failure occurs depends on loading rate and stress distribution.
Each intervertebral disk has a nucleus pulposus surrounded by a thick capsule called the annulus fibrosus (Figure 1–21). End-plates composed of hyaline cartilage separate the intervertebral disk from the vertebral body. The unique interplay of the nucleus pulposus, annulus fibrosus, and end-plates accounts for the ability of the disk to withstand compressive, rotational, and shear forces. It is likely that hedgehog genes and the bone morphogenetic protein (BMP) inhibitors, including Pax-1, sonic hedgehog, Indian hedgehog, and Noggin genes, are key factors involved with intervertebral disc formation.
The nucleus pulposus lies in the center of the intervertebral disk, except in the lumbar spine, where it lies slightly posterior, at the junction of the middle and posterior thirds of the sagittal diameter. The nucleus pulposus is composed of a loose network of fine fibrous strands in a gelatinous matrix that contains water-binding glycosaminoglycans. The number of glycosaminoglycans decreases with age, thereby decreasing the hydration of the nucleus pulposus.
The annulus fibrosus is the ringlike outer portion of the disk and consists of fibrocartilage and fibrous tissue. The fibrocartilage is in a series of concentric laminated bands. In the first band, the collagen fibers are principally oriented at a 30-degree angle in one direction; in the second band, they are oriented at a 30-degree angle in the opposite direction; and the pattern continues (Figure 1–21A), with the result that the annular fibers form an intricate crisscross arrangement (Figure 1–21B). Centrally, the collagen fibers of the annulus fibrosus are attached to the cartilaginous end plates. Peripherally, the fibers are attached to the bone of the vertebral body by Sharpey fibers.
The interaction between the nucleus pulposus and the annulus fibrosus accounts for the mechanical behavior of the intervertebral disk. The mechanical properties of the disk are viscoelastic and therefore depend on the loading rate and duration.
During compressive loading, the stress is transferred from the vertebral end plates to the intervertebral disk. With compression, pressure increases in the nucleus pulposus, and the fluid exerts hydrostatic pressure on the annulus fibrosus. As a result, the central portion of the vertebral end plates are pushed away from one another, and the annular bands are pushed radially outward. The bulging annular bands develop tensile stress in all directions, the optimal orientation for maximal mechanical strength for the collagen fibers.
When the nucleus pulposus ages, its hydration decreases and its hydrostatic properties change. The load-transferring mechanism of the disk is greatly altered if sufficient hydrostatic pressure does not develop. In this situation, the annulus fibrosus transfers the stress to the periphery of the intervertebral disk; however, the fibers are subjected to compressive stress, which is not the optimal loading orientation for collagen fibers. This situation could lead to inadequate stress transfer from successive vertebral bodies, and this in turn could result in compression fractures of the vertebral bodies.
The nucleus pulposus has no effect during tensile loading of the intervertebral disk. Tensile loads are supported by tensile and shear stresses in the annulus fibrosus. The orientation of the collagen fibers of the annulus fibrosus provides no ability to resist shear stresses. Therefore, disk failure is greater with tensile loading than with compressive loading. Excessive shear stresses in the intervertebral disk may cause failure in pure rotational loading when the nucleus pulposus has insufficient load to apply its hydrostatic effects to the annulus fibrosus.
Pain located in the low back or neck, without radiation to the extremities, is frequently attributed to the structures in the area, namely the intervertebral disk and/or the facet joints. The relief of pain achieved by joint replacement in other areas of the body has provided an impetus to replace the joints in the spine. Tight regulation of these interventions has resulted in a delay in getting these types of prostheses to the U.S. market. Other countries are actively marketing prostheses for the cervical disc and the lumbar disc. Several designs for each area are now in clinical studies in the United States. In addition to pain relief, these prostheses offer the maintenance of at least some motion in the disc space, compared to fusion. Some of these replacements are metal on metal; others are designed like a traditional joint replacement with a polyethylene bearing component. The efficacy of these devices remains to be seen, especially when compared to the potential complications that might ensue.
Peripheral nerves are heterogeneous composite structures comprised of multiple cell types including neurons, Schwann cells, macrophages, and fibroblasts. The primary function of peripheral nerves is the propagation of action potentials. To improve the speed of action potential propagation (ie, nerve conduction velocity) without increasing the diameter of the axon, the Schwann cell, the primary glial cell of the peripheral nervous system, forms a myelin sheath around the axon to insulate the axon and reduce dissipation of the action potential.
The peripheral nerve is composed of motor, sensory, and sympathetic nerve fibers. Such fibers can be either myelinated or unmyelinated. The average diameter of myelinated fibers is 2–22 m, and unmyelinated axons are 0.4–1.25 m. Surrounding the individual myelinated nerve fibers or the group of unmyelinated nerve fibers is the endoneurium, a collection of thin collagen strands that provides adequate nourishment and protects the individual axons. Multiple nerve fibers collect to form a group of fibers called afascicle. Such fascicles are bound and encircled by the perineurium, a collection of a connective tissue sheath composed of perineurial cells. The perineurium is the major contributor to the nerve's tensile integrity and strength and is also the blood–nerve barrier. Several fascicles may be arranged into a group fascicle surrounded by connective tissue, termed the internal or interfascicular epineurium. The connective tissue that surrounds the periphery of the entire nerve is called the external or extrafascicular epineurium. The primary function of this connective tissue structure is to nourish and protect the fascicles. Fascicular patterns are divided into the following three types: monofascicular, oligofascicular, and polyfascicular. Monofascicular patterns consist of one large fascicle, whereas oligofascicular patterns consist of a few fascicles; polyfascicular patterns consist of many fascicles of varying sizes that can be arranged with or without groupings of fascicles. Nerves found in the upper arm are routinely polyfascicular. In its course from the upper arm to the fingertips, a peripheral nerve undergoes changes from polyfascicular pattern in the upper arm, oligofascicular in the elbow region, and monofascicular in the hand and fingers. For example, the ulnar nerve is polyfascicular as it exits the brachial plexus until just before the elbow, at which point it becomes oligofascicular. After the division into the motor branch at the wrist, the pattern is monofascicular. These patterns may help determine which type of nerve repair is appropriate for a particular nerve injury. In surgical nerve repair, proper identification of fascicular arrangement is crucial to achieving a successful outcome. Peripheral nerves are extensively vascularized with separate yet interconnected microvascular systems in the epineurium, perineurium, and endoneurium. The vascular pattern of the peripheral nerve is characterized by longitudinally oriented groups of vessels, with a great number of communicating anastomoses. The vasculature is composed of an intrinsic vascular system consisting of vascular plexa in the epineurium, perineurium, and endoneurium and an extrinsic system derived from closely associated vessels running with the nerve. From a surgical perspective, the role of the intraneural microvascular system is vital in regard to the effects of chronic irritation, compression, mobilization, stretching, and transection. If considering a group fascicular or fascicular repair technique, the effects of intraneural vascular damage secondary to surgical manipulation need to be considered.
ROLE OF MYELIN SHEATH
One of the primary functions of a myelin sheath is to provide faster, more efficient transmission of the nerve impulse. Myelin itself acts as an insulator around the axon, reducing the dissipation of the action potential into the surrounding medium as it propagates through the axon. There are discontinuities in the myelin sheath along the nerve fiber that are known as nodes of Ranvier and contain high concentrations of voltage-gated sodium channels. The arrival of an action potential at a node depolarizes the membrane, opening the sodium channels to induce a massive influx of sodium ions into the axon. This generates an electrical pulse that is propagated down the axon to the next node of Ranvier, a process termed saltatory conduction. Saltatory conduction allows not only for an increased speed of nerve impulse transmission, but also reduces the energy requirement as well. Myelin is not just a conduit for conduction, however. Several myelinopathies, including type I Charcot-Marie-Tooth disease and multiple sclerosis, feature axonal pathology secondary to myelin dysfunction. Although the reciprocal relationship between neurons and glial cells is maintained, many of the clinical manifestations of these diseases are directly related to secondary axonal loss rather than to primary myelin dysfunction. Furthermore, mice with mutations in genes that encode the myelin-specific proteins myelin-associated glycoprotein (MAG), proteolipid protein (PLP), and ciliary neurotrophic factor (CNTF) exhibit evidence of axonal degeneration. MAG localizes to the periaxonal membrane and has a similar structure to members of the immunoglobulin family that participate in cell adhesion. MAG-deficient mice seem to develop apparently normal myelin sheaths and appear phenotypically normal. However, the caliber of myelinated axons decreases, and by 8 months functional contact between the axon and myelin is disrupted and the axons undergo a degenerative process similar to many human peripheral neuropathies. PLP is the major protein constituent of central nervous system myelin. PLP-deficient mice initially appear phenotypically normal and exhibit only minor myelin abnormalities, but they eventually develop behavioral motor abnormalities resulting from ongoing swelling and degeneration of small-diameter myelinated axons. CNTF is a neurotrophic factor that plays an important role in the neural response to injury. Adult CNTF-null mutant mice exhibit a progressive atrophy and degeneration of large-diameter motor neurons, in addition to disruption of the axon–Schwann cell communication network in the paranodal region, resulting in a small reduction in muscle strength. This evidence suggests a functional role for myelin in regulating the maturation, maintenance, and viability of myelinated axons.
Myelin is formed from the extension of the plasma membrane of Schwann cells. Myelin is unique in its high lipid-to-protein ratio, including high contents of galactosphingolipids and saturated long-chain fatty acids. For myelin to achieve its insulating properties, nonconducting structures as well as aqueous cytosolic molecules must be removed. Myelin excludes much of its conductive extracellular and cytoplasmic material by fusing large surfaces of the cytoplasmic leaflets of its plasma membrane, forming a morphological structure called the major dense line. The apposing exoplasmic surfaces of the plasma membrane also fuse to form intraperiod lines. Schmidt-Lanterman incisures are cytoplasmic channels that extend from the internal limit to the external limit of the major dense line. These channels facilitate communication between the axon and Schwann cell body and assure adequate nutritional supply within the regions of compacted myelin.
IMPLANT MATERIALS IN ORTHOPEDICS
The body is a harsh chemical environment for foreign materials. An implanted material can have its mechanical and biologic properties significantly altered by body fluids. Degradation mechanisms, such as corrosion or leaching, can be accelerated by ion concentrations and pH changes in body fluids. The body's response to an implant can range from a benign to a chronic inflammatory reaction, with the degree of biologic response largely dependent on the implanted material. For optimal performance in physiologic environments, implant materials should have suitable mechanical strength, biocompatibility, and structural biostability. As the field of biomaterials science developed, various classification schemes for implantable materials were proposed, including schemes based on chemical composition and biologic response.
Implant materials can be classified as biotolerant, bioinert, and bioactive. Biotolerant materials, such as stainless steel and PMMA, are usually characterized by a thin fibrous tissue layer along the bone-implant interface. The fibrous tissue layer develops in part as a result of leaching processes that produce chemicals that irritate the surrounding tissues. Bioinert materials, such as cobalt-based alloys, titanium, and aluminum oxide, are characterized by direct bone contact, or osseointegration, at the interface under favorable mechanical conditions. Osseointegration is achieved because the material surface is chemically nonreactive to the surrounding tissues and body fluids. Bioactive materials, such as calcium phosphate ceramics, particularly hydroxyapatite, have a bone-implant interface characterized by direct chemical bonding of the implant with surrounding bone. This chemical bond is believed to be caused by the presence of free calcium and phosphate ion groups at the implant surface. The calcium phosphate materials can be used as implants or coatings. Other bioactive materials are the growth factors, which are finding application in stimulating desired responses from connective tissues.
Minimizing the local and systemic response to an implanted material through improved biocompatibility is only one engineering concern for reconstructive implant surgery. A prosthetic implant must appropriately transfer stress at the bone-implant surface to ensure long-term implant stability. Nonphysiologic stress transfer may cause pressure necrosis or resorption at the bone-implant interface. Necrotic and resorbed bone may lead to implant loosening and migration, thus compromising implant longevity. Polyethylene wear particles are linked to osteolysis, also compromising implant longevity. Moreover, it is essential that materials have properties capable of sustaining the cyclic forces to which the implant will be subjected. For example, if the material properties are not adequate for load sharing, the implant may fail because of fracture. If the geometry and material properties of the implant make it too rigid in comparison with the bone, then stress shielding of the bone is likely to occur, making bone resorption and implant loosening inevitable.
In addition to acceptable biocompatibility characteristics, biomaterials must demonstrate material properties suitable for their desired use. Materials used to manufacture total joint replacement systems must demonstrate a yield stress that is greater than the stress expected from joint forces but must also have a flexural rigidity that will not result in unacceptable amounts of stress shielding of the bone. General stress–strain curves for the classes of materials allow for the comparison of material properties (Figure 1–22). For instance, ceramics are characterized by a high elastic modulus but are extremely brittle. In contrast with ceramics, metals have a lower elastic modulus but demonstrate increased ductility.
The most commonly used biomaterial combinations for orthopedic joint replacement are metals and metal alloys articulating with ultrahigh-molecular-weight polyethylene (UHMWPE). Stainless steel, an iron-based alloy, was used in Charnley's original hip prosthesis and is the material most commonly used for internal fixation plates, rods, and screws. Advances in materials science have produced stronger cobalt-based and titanium-based alloys. The wear resistance of cobalt-based alloys makes them desirable for applications involving articulating surfaces. Titanium-based alloys, which have a modulus of elasticity closer to that of bone than the other metal alloys do, are currently being manufactured as femoral hip stems to reduce the effects of stress shielding.
Polymers and ceramics are also important classes of materials for orthopedic implant applications. UHMWPE has a low coefficient of friction, making it ideal for an articulating surface. PMMA is used as a grouting agent in total joint arthroplasty to provide immediate fixation of total joint components to the skeleton. Porous-coated components require ingrowth of tissue to the porous coating over a period of weeks or months to obtain stability. Aluminum oxide and zirconium oxide have gained popularity as materials for ceramic femoral heads because of their high wear resistance and low coefficient of friction. Finally, calcium phosphate ceramics, particularly hydroxyapatite, are used in monolithic form as an augmentation material for metaphyseal bone defects and as a coating on metal devices for total joint arthroplasty.
The suitability of a metal component for maintaining longevity of a total joint replacement depends on the design of the implant and the biocompatibility, strength, wear, and corrosion characteristics of the metal. Materials scientists can improve on one or several of the characteristics of a metal alloy by varying the composition or by using different manufacturing processes.
An understanding of the terminology used to describe the strength and stiffness characteristics of a metal is essential in making informed decisions about the different metal alloys. The most important characteristics are elastic modulus, yield stress, ultimate tensile stress, and fatigue stress. As discussed at the beginning of this chapter, these properties can be determined from stress–strain curves and fatigue curves. The composition specifications and mechanical characteristics of all metals and their alloys used for orthopedic implants were standardized by the American Society for Testing and Materials (ASTM).
The grain size, inclusion content, and surface porosity influence the strength characteristics of a metal. In general, the larger the grain size, the lower the tensile and fatigue strength at fracture. Excessive inclusions or a high surface porosity weakens the metal by acting as stress risers and by providing areas for crevice corrosion. Manufacturing processes can be used to control these factors. For example, heating metal to a temperature near its melting point increases the grain size, whereas forging processes decreases the grain size.
Corrosion is a chemical reaction process that weakens the metal. Three types of corrosion are prevalent in implant materials: fatigue, galvanic, and crevice corrosion. Although all metals corrode in the physiologic environment, the severity of corrosion is determined by the chemical composition of the metal. Stainless steel corrodes more readily than either cobalt-based or titanium-based alloys. The chromium and molybdenum content of both stainless steel and cobalt-based alloys produces a corrosion-resistant surface layer. Titanium-based alloys have an adherent oxide passive film layer that provides their corrosion resistance.
The surfaces of all metallic implants are passivated (made passive to corrosion) with nitric acid to form an oxide surface layer that increases corrosion resistance. Fatigue corrosion may occur, however, if this passive film layer on the implant surface was scratched or cracked and does not self-passivate in vivo. The ability to self-passivate may be hindered by wear processes or micromovement between modular components, a process called fretting. Once corrosion begins, the implant weakens and will fail at a stress level below the endurance limit of the metal.
Galvanic corrosion occurs when an electric current is established between two metals that have different chemical or metallurgic compositions. Some differences arise from manufacturing processes and may be subtle, as in the difference between annealed bone plates and cold-worked screws made of stainless steel. Other differences that lead to galvanic corrosion arise from the close contact of two different metals in an implant, such as a titanium alloy femoral stem in contact with a cobalt alloy head. An evaluation of retrieved mixed metal femoral hip components consisting of a cobalt-chromium modular head on a titanium alloy stem demonstrated some degree of corrosion in the majority of the components. Further evaluation determined that corrosion occurred in all components that were implanted for longer than 40 months. The long-term clinical significance of the presence of corrosion caused by femoral component modularity is unknown. To avoid catastrophic galvanic corrosion, however, stainless steels should never be used with either cobalt-based or titanium-based alloys.
Crevice corrosion generally occurs when the fluid in contact with a metal becomes stagnant, resulting in a local oxygen depletion and a subsequent decrease in pH in relation to the rest of the implant. This form of corrosion is most prevalent underneath bone plates at the screw-plate junction. The mechanism of crevice corrosion, however, is apparent in point or structural defects in a metal. Corrosion of a defect results in progressive deepening of the defect, leading to the development of large stress concentrations and catastrophic failure of the implant.
There are four major groups of iron-based alloys or stainless steels, classified according to their microstructure. The group III (austenitic) stainless steels, which are labeled 316 and 316L, are used for orthopedic implants. The difference between 316 and 316L is that the latter contains a smaller percentage of carbon. Lowering the carbon content increases the corrosion resistance. Among the various elements contained in 316 and 316L stainless steels is molybdenum, which hardens the passive layer and increases pitting corrosion resistance.
Iron-based alloys have a wide range of mechanical properties (Table 1–3) that make them desirable for implant applications. Despite composition modifications, stainless steels are susceptible to corrosion inside the body, however. Therefore, they are most appropriate for temporary devices such as bone plates, bone screws, hip nails, and intramedullary nails.
Corrosion of stainless steels occurs for one of several reasons. The most common reason is incorrect metal composition, which increases the chance that galvanic corrosion processes will occur. Molybdenum is added to these metals to increase corrosion resistance; however, too much molybdenum can embrittle the alloy. Chromium carbide may form between the grain boundaries and result in grain boundary corrosion. This phenomenon is referred to as sensitization.
Another reason for corrosion is mismatch of implant components, especially when bone plates and screws are used, because even implants manufactured by the same company in different lots can be susceptible to corrosion processes caused by compositional differences. Crevice corrosion can occur at the junction of the screw with the bone plate and develops from local changes in pH and oxygen concentration that may result from slightly different manufacturing processes of the components.
Leaving plates and screws used to fix fractures in younger patients increases the risk of slow progressive corrosion over the years. Failure of the plate resulting from corrosion processes may also lead to bone fracture because stress shielding invariably occurs under the plate. Titanium alloy plates and rods are gaining popularity because of their corrosion resistance and lower elastic modulus, properties that lower the degree of stress shielding.
The mechanical properties that make cobalt-based alloys suitable for load-bearing implant applications are summarized in Table 1–3. Among the elements contained in these alloys is molybdenum, which is added to produce finer grains and thereby results in higher strength. The cobalt-based alloys are characterized by high fatigue resistance and high ultimate tensile strength levels, properties that make them appropriate for applications requiring a long service life and ability to resist fracture. The high wear resistance of these alloys also makes them desirable for load-bearing and articulating surface applications. Cobalt-chromium alloys are primarily used for components in total joint implants.
Despite the advantages of cobalt-based alloys, some reported cases show that surface porosities act as stress risers and lead to premature fatigue failure. Hot isostatic pressing (HIP)—a process that involves simultaneously applying both heat and pressure to consolidate powder into a solid form—was adapted to significantly reduce surface porosity in cast metals. After this process is performed, the material must be heat treated to attain maximal benefit. When performed properly, HIP increases the fatigue resistance, static strength characteristics, and corrosion resistance of cobalt-based alloys. Unfortunately, the heat treatment necessary to apply porous coatings, such as beads, obviates much of the benefit of HIPing.
TITANIUM AND ZIRCONIUM ALLOYS
Commercially pure titanium and titanium-based alloys are metals of low density (4.5 g/cm3) and have chemical properties suitable for implant applications. Zirconium is in the same group in the periodic table. Both metals form an adherent, highly stable oxide surface layer that makes them extremely resistant to corrosion and chemically nonreactive to the surrounding tissues.
The mechanical properties of titanium and titanium-based alloys are summarized in Table 1–3. The elastic modulus value for titanium-based alloys is approximately 110 GPa, which is approximately half the value for iron-based or cobalt-based alloys but is still at least five times greater than the value for bone. Zirconium, when alloyed with niobium (2.5%) and oxygen (0.11%), has a modulus of 97.9 GPa. The higher the impurity content of the metal, the higher the strength and brittleness. Titanium and zirconium can be alloyed together and in combination with a variety of other metals, including niobium, tantalum, molybdenum, and iron. Because of their low density, these alloys have superior specific strength (strength per density) over all other metals. Titanium has poor shear strength and wear resistance, however, making it unsuitable for applications involving articulating surfaces. It also exhibits notch sensitivity, which means that a small flaw or crack on the surface, such as might occur with mechanical damage, can cause a tremendous reduction in strength and increase the susceptibility to fracture.
New manufacturing techniques are attempting to improve titanium-based alloys as bearing surfaces. Nitrogen ion implantation techniques were evaluated for their ability to increase the wear resistance and surface hardness of the alloys. The ion implantation process embeds elemental nitrogen ions in the surface, which causes distortions or strains within the crystal lattice of the metal and results in increased surface microhardness. The results are not adequate to allow titanium or its alloys to be used as bearing surfaces. However, an oxidized zirconium alloy (ASTM F2384) has produced a bearing surface that is showing promise in clinical trials for the last several years. In vitro studies show that the increased surface hardness from the oxide layer significantly improves the wear resistance of the treated implant. The surface oxide layer is very thin (approximately 0.004 mm) but extremely adherent.
Polymers have a wide range of properties attributable to variations in their chemical composition, structure, and manufacturing process, which make them suitable for several different implant applications. The choice of polymer for application is dictated by the effect of the physiologic environment on the stability of the material. Some polymers, such as PMMA, leach toxic substances into the surrounding tissues. Conversely, other polymers, such as silicone, absorb fluids from the body, and this absorption alters the mechanical properties. Despite the possible consequences of polymer implantation, the use of polymers as implant materials has been successful.
All polymers are composed of long chains of repeating units. These units may form linear, cross-linked, or branched chains. The individual chains may be organized in an orderly crystalline form having parallel or folded chains, or they may have an amorphous structure or a mixed structure. The molecular weight, chemical composition, degree of crystallinity, size and polarity of side groups, and degree of cross-linking determine the mechanical properties of the polymer. In general, as the molecular weight and crystallinity increase, the tensile strength and the resistance to cracking increase. The crystallinity is decreased by copolymerization, branching of chains, and large side groups.
UHMWPE possesses an array of properties, including high abrasion resistance, low friction, high impact strength, excellent toughness, low density, ease of fabrication, biocompatibility, and biostability, that makes it an attractive material for use in fabricating bearing surfaces for total joint replacements. UHMWPE is the material of choice for the liner of acetabular cups in total hip arthroplasties, the tibial insert, patellar components in total knee arthroplasties, and other load-bearing interfaces. The clinical performance of these components for nearly three decades has been excellent; however, there are concerns about the long-term wear of these devices. The concern is not only that the materials will wear out, but also that the wear debris generated will evoke an undesirable biologic reaction. It is known that particles in the submicron size undergo phagocytosis, resulting in a variety of biologic reactions. The reactions can include granulomatous lesions, osteolysis, and bone resorption. The phenomenon of wear of UHMWPE in total joint replacement is widely regarded as one of the most challenging problems in contemporary orthopedics.
A variety of factors influence the wear of polyethylene. These include the nature or quality of the starting material; the degree of cross-linking, strength, and toughness of the material; the manufacturing technique; the thickness of the polyethylene; the component sterilization conditions; and the storage environment and age of the component. The composition and physical properties of the starting materials used in the fabrication of UHMWPE components—in particular, molecular weight—have a profound effect on the material's performance. It is widely agreed that low-molecular-weight impurities and a high crystallinity are detrimental to the clinical performance. Increased crystallinity causes less resistance to crack initiation, crack propagation, and oxidation-enhanced wear.
UHMWPE is either machined from ram-extruded bar stock or compression molded directly from the starting powder. Ram extrusion is a two-step process in which the starting powder is first heated and pressurized in a polymerizing chamber followed by extrusion of bar stock. The second step, usually performed by the implant manufacturer, is the machining of the component with a precision cutting tool. By contrast, direct compression molding is a one-step process. A preshaped component and premeasured volume of starting powder are placed into a heated mold, and the component is then formed under pressure. Each manufacturing process has its own unique characteristics and problems. Some engineers believe that compression-molded components, although more expensive to produce, result in products with enhanced in vivo wear performance.
In general, as the thickness of the UHMWPE components decreases, the stresses increases, and therefore, the wear also increases. This is because with an increase in thickness, the structural stiffness of the component increases, even when the value of the modulus of elasticity of the material remains unchanged. This effect has resulted in a recommendation for the minimal thickness of acetabular liners of approximately 5–6 mm and tibial inserts of 7–8 mm.
Several manufacturers choose gamma irradiation for sterilizing UHMWPE components at doses between 25,000 and 40,000 Gy. Gamma radiation causes both cross-linking and chain fission and is the major contributor to subsequent surface and subsurface oxidative degradation of the components. Other manufacturers choose surface sterilization techniques, such as ethylene oxide, or one of the several plasma techniques, which have no effect on the bulk properties of the polyethylene. There is considerable literature on the effect of a number of variables associated with sterilization, packaging, and aging of polyethylene on the in vivo and in vitro physical and mechanical properties. These include the method of sterilization, sterilization dosage, packaging atmosphere, and aging conditions. For example, irradiation and storage in an inert atmosphere reduces the oxidative degradation before implantation. Because of this volume of work, no consensus has yet emerged on these effects. As noted, gamma sterilization, particularly in air, causes significant reduction of the various mechanical characteristics that are related to the subsequent wear and performance of the material. These parameters include crystallinity, melting temperature, oxidation strength, tensile properties, and density.
Irradiation of polyethylene at much higher doses (50,000–150,000 Gy) followed by an annealing process results in a cross-linked polyethylene with different properties. Early hip simulator wear studies indicate that this cross-linked polyethylene has very low wear rates and superior wear properties compared with conventional polyethylene. Early clinical trials seem to confirm a reduction in wear rates but not to the magnitude predicted by the wear simulator studies.
Self-curing PMMA, commonly used as a grouting agent, is often called the weak link in total joint arthroplasty. Compared with cortical bone, bone cement has a lower elastic modulus and significantly inferior mechanical strength properties. The tensile strength of PMMA is similar to that of cancellous bone. The low modulus of elasticity allows for gradual transfer of stress from implant to bone. Mechanically, PMMA is weakest in shear loading and strongest in compression loading configurations.
Implant design and cementing technique must compensate for weakness in tension, to avoid catastrophic failure of the cement. The poor fatigue strength of PMMA can be attributed primarily to porosity. Studies show that PMMA porosity is increased and fatigue strength is further decreased by mixing the cement with chilled monomer, rather than with monomer at room temperature. Therefore, if chilled monomer must be used, it is crucial to use porosity reduction techniques, such as centrifugation or vacuum mixing, concurrently. The fatigue strength of PMMA is also decreased by the presence of inclusions, such as bone chips and blood.
Aside from the inherent mechanical weakness of PMMA, the polymerization process causes local and systemic biologic effects. Locally, adjacent tissues can become necrotic because of the extreme heat of polymerization, which can generate temperatures approaching 100° C. Systemically, the leaching of monomer during the curing process may cause hypotension.
The fatigue properties of PMMA manufactured by different companies vary because of intrinsic compositional differences, such as the size of the polymer beads, the addition of copolymers, and the presence of additives and radiopacifiers. In a study of the fatigue life of five commonly used bone cements (CMW, LVC, Palacos R, Simplex P, and Zimmer Regular), investigators prepared each cement in the manner suggested by its manufacturer. They found that Palacos R and Simplex P had equivalent fatigue strengths and that these two products had significantly greater fatigue strengths than the other three products. For each product, investigators compared the fatigue life of a regular sample with that of a sample that had undergone a process to reduce its porosity. In the case of each product, a reduction in the cement's porosity increased its fatigue life. Moreover, when investigators centrifuged two packages of Simplex P mixed with chilled monomer for 60 s, they found a fivefold increase in the fatigue properties of this cement.
Antibiotics were added to bone cement at the time of surgery for years to reduce the risk of infection or to treat infection. Heat-stable antibiotics that are eluted from the bone cement inhibit antibiotic growth for variable periods from days to weeks. To obtain more uniform formulations, manufacturers have now obtained permission from the Food and Drug Administration (FDA) to make the bone cement with premixed antibiotics. Studies show that the fatigue strength of the cement is not reduced with these additions.
Ceramics are wear resistant and strong in compression, but they are extremely brittle and susceptible to cracking. Ceramic materials must be carefully chosen for specific implant applications because chemical composition affects the mechanical properties and biologic responses of each ceramic. For instance, in calcium phosphate ceramics, an alteration in the ratio of calcium to phosphorous can significantly alter the in vivo dissolution rate of the ceramic.
The mechanical properties of ceramics depend on grain size, porosity, density, and crystallinity. Strength is normally improved with increased density, increased crystallinity, and decreased porosity. The hardness and wettability of ceramics and the fact that ceramics can be polished to smooth surfaces make them ideal candidates for bearing surfaces. Nevertheless, for a ceramic implant to be reliable, its design must avoid sharp corners and notches, to overcome the predictable mechanical flaws of the material.
The catastrophic effects of implant loosening associated with polyethylene wear debris led to interest in using other materials at the articulating surface. The use of aluminum oxide (Al2O3) was explored because it is a highly biocompatible material with high frictional resistance. In fact, the coefficient of friction for alumina-on-alumina articulations is approximately 2.3 times less than the coefficient for metal-on-polyethylene articulations. Studies show that alumina-on-alumina articulations demonstrate approximately 5000 times less wear than metal-on-polyethylene articulations do under experimental loading conditions.
In clinical practice, all-alumina acetabular components have not performed well, probably because of loosening caused by the modulus mismatch between aluminum oxide and bone. Designs to circumvent this problem have fixed an alumina bearing component inside a metal ingrowth cup, so the bone can grow into a porous metal surface. Ceramic-on-polyethylene articulations show clinical promise, however. Aluminum oxide has excellent wear characteristics, and any ceramic wear debris that does accumulate at the interface may be less bioreactive than polyethylene or PMMA wear debris. Alumina-on-alumina articulations demonstrate very low wear rates when clinically applied with a metal shell for attachment to acetabular bone. Although the alumina wear rates are low, problems have occurred because of placement of the components, allowing impingement and abrasion of the neck of the femoral component.
Despite the excellent wear and friction characteristics of aluminum oxide, its fracture toughness and tensile strength are relatively low. The fracture rate of alumina is low but significant. In one series, the authors reported 13 failures out of 5500 implantations over 25 years (0.23%), and disturbingly, they were unable to attribute some of the fractures to any rational explanation. Fracture can be a result of microstructural flaws, which are statistical in nature. The elastic modulus of aluminum oxide is approximately 20 times greater than that of cortical bone. Increased grain size decreases strength and a large grain size is linked to reported cases of catastrophic wear. Careful regulation of manufacturing processes is necessary to obtain reliable aluminum oxide implants with small grain size, high density, high purity, adequate strength, and adequate component size. With these caveats, these alumina-on-alumina implants may have a role in hip replacement in younger patients because of their reduced rate of particulate wear production. The high strength and hardness of alumina also recommend it for use with polyethylene bearings.
Zirconium oxide (Zirconia) temporarily became an attractive material for highly loaded joint replacement applications. Pure zirconium oxide can be maintained in a metastable, tetragonal crystal structure with fine grain structure, with the addition of a stabilizing oxide, such as yttrium oxide (Y2O3).
In comparison with aluminum oxide, zirconium oxide exhibits increased fracture toughness, increased bending strength, and decreased elastic modulus. Moreover, Zirconia-to-polyethylene articulations demonstrate favorable wear characteristics compared with alumina-to-polyethylene articulations in vitro. Because the mechanical and wear properties of zirconium oxide are superior to those of aluminum oxide, it is possible to manufacture smaller femoral heads for low-friction total hip arthroplasty. However, several studies show an increase in the monoclinic phase of Zirconia over time in vivo, calling into question the long-term biostability of the material. Further, the phase change was associated with an increase in surface roughness and wear rates of polyethylene bearing surfaces. Other problems have plagued proponents of Zirconia as a bearing material. Catastrophic failure was seen in a high proportion of Zirconia heads produced in a tunnel kiln in Europe, which were shown on retrieval to have high (21–68%) monoclinic phase present. Zirconia femoral heads may need significant research to demonstrate safety and regain consumer confidence. Although still available in some markets, Zirconia femoral heads may have a very limited place in the orthopedic market.
Calcium phosphate ceramics, classified as polycrystalline ceramics, have a material structure derived from individual crystals that become fused at the grain boundaries during high-temperature sintering processes.
Tribasic calcium phosphate [Ca10(PO4)6(OH)2], which is commonly called hydroxyapatite, is a geologic mineral that closely resembles the natural mineral in vertebrate bone tissue. Tribasic calcium phosphate should not be confused with other calcium phosphate ceramics, especially tricalcium phosphate [Ca3(PO4)2], which is chemically similar to hydroxyapatite but is not a natural bone mineral.
Bulk hydroxyapatite (HA) is manufactured from a starting powder, and the manufacturing process consists of compression molding and subsequent sintering. Macroporous ceramics can be obtained by combining the starting mixture with hydrogen peroxide. Otherwise, a dense structure with a small percentage of micropores results. Dense HA ceramics have a compressive strength greater than that of cortical bone; however, their tensile strength is approximately 2.5 times less than their compressive strength. Small reductions in density can significantly reduce tensile characteristics of the ceramic. Bulk HA can also be formed with a very uniform pore structure by the chemical conversion of calcium carbonate coral structures. This material can be used to augment bone graft in cancellous bone areas.
Although the static mechanical properties of bulk HA are good, the resistance to fatigue failure is low in physiologic conditions, as is common with sintered ceramics and particularly bioactive ceramics. Therefore, bulk HA is not suitable for applications requiring mechanical loading. Bulk HA is used successfully in clinical practice as a bone graft substitute to fill defects associated with fractures or as a bone graft extender for fractures or fusions. A composite prosthesis made by plasma spraying thin (approximately 50 m) coatings of HA or other calcium phosphate compounds onto a metal substrate was developed and is able to withstand the physiologic stresses imposed on it while providing an osteoconductive surface to achieve optimal bone apposition and ingrowth. These coatings can be applied by other means including chemical deposition from solution. Generally, HA coatings do not provide adequate fixation of bone when applied to smooth metal surfaces, and they require a roughened or porous surface for long-term fixation. Experimental results indicate that HA stimulates more extensive and uniform growth of bone into a porous-surfaced femoral stem or acetabular cup and probably aids bone growth across gaps between implants and surrounding bone. The durability of the fixation of the prosthesis to bone may be related to the quality and durability of the coating. HA coatings on ingrowth/ongrowth surfaces is probably not needed in the primary procedure but is considered by many surgeons to be helpful in the revision procedure, to obtain more reliable bone ingrowth.
Several formulations of an injectable hydroxyapatite are now available for minimally invasive surgery. These materials are similar to PMMA in that the setting process is a chemical reaction that yields the final product, in this case hydroxyapatite, or a similar calcium phosphate compound.
Carbon occurs naturally in many forms, each having different structures, material properties, and uses. Coal is an example of carbon in its amorphous form with no crystalline structure. Graphite has an organized crystalline structure, in which carbon atoms are arranged in two-dimensional hexagonal sheets tightly joined by strong covalent bonds. Diamond is a third form of carbon having a three-dimensional cubic crystal structure with increased atomic bonds.
Pyrolytic carbon is a manufactured material formed by the pyrolysis, or heating, of gaseous hydrocarbons, causing them to decompose to a stable gas and carbon. The resulting pyrolytic carbon is usually deposited on a graphite substrate in a turbostratic, two-dimensional crystalline structure with a high concentration of three-dimensional diamond cross-link bonding. In general, the physical properties of pyrolytic carbon fall between those of graphite and diamond. The form of pyrolytic carbon used as a surgical implant material has a very fine grain structure (approximately 50 Å), and isotropic physical and mechanical properties. Pyrolytic carbon is chemically inert and resistant to wear and mechanical fatigue with a fatigue strength at 109 cycles close to its failure strength. It also has a high fracture strength, and low elasticity modulus (21–26 GPa), which falls within the range of moduli reported for cortical bone. Its biocompatibility is well documented and has been confirmed clinically in extensive use in cardiovascular implants for more than 30 years. Pyrolytic carbon has also proved to be extremely biocompatible in both osseous and soft tissues.
Pyrolytic carbon was evaluated in cardiovascular, dental, soft-tissue, and orthopedic implants. It is the material of choice for the construction of mechanical artificial heart valves. The human heart beats on an average of 100,000 times per day, or 35 million times per year. The carbon-on-carbon pivot systems in heart valves must resist stress and wear each time the heart beats. The history of successful function in heart valve components demonstrates the outstanding wear resistance of the carbon-on-carbon articulation, biocompatibility, high fatigue strength, and structural durability of the pyrolytic carbon material. Pyrolytic carbons were evaluated clinically in orthopedics for replacement of the small bones and joints of the hands and feet. Nonconstrained pyrolytic carbon metacarpophalangeal joint replacements were evaluated in human clinical trials with results at long-term follow-up demonstrating excellent performance of the implants with a 15-year survival rate of more than 70%. The long-term clinical results of the metacarpophalangeal joint arthroplasties verify the biologic and biomechanical compatibility of pyrolytic carbons in a demanding orthopedic application. With FDA approval, these joints are now being marketed for general use.
Growth factors hold great promise for the treatment of a variety of musculoskeletal conditions. The response of bone to structural damage is nearly unique in biology. The vast majority of tissues, when traumatized, heal with a fibrous scar, the cells and structure of which are not normal and are unable to assume fully the function of the tissue. In contrast, bone, the cornea, and the liver are capable of true cellular, morphologic, and functional regeneration. The initial phase of bone healing is characterized by an inflammatory response in consolidation of the hematoma within the fracture site. This is followed by the proliferation of periosteal, endosteal, and marrow stromal cells adjacent to the site, and recruitment of undifferentiated mesenchymal cells from nearby soft tissues. These cells and their progeny differentiate to become chondroblasts, chondrocytes, osteoblasts, and osteocytes. The cartilage formed is eventually replaced by bone, and the early woven bone is remodeled to a more mature lamellar structure.
Growth factors are polypeptides that serve as signaling agents for cells. These local proteins bind to specific receptors, occasionally with the assistance of extracellular binding proteins, to stimulate or inhibit functions inside the cell during development and throughout an organism's life. For example, growth/differentiation factor-5 (GDF-5) is essential for normal appendicular skeletal and joint development in humans. GDF-5 acts at two stages of skeletal development and by two distinct mechanisms: (1) promoting the initial stages of chondrogenesis by promoting cell adhesion and (2) increasing the size of the skeletal elements by increasing proliferation within the epiphyseal cartilage adjacent to its expression within the joint interzone. The discovery of these substances revolutionized the field of cell biology by revealing the mechanisms of regulation of cell activities. Growth factors are present in plasma or tissues, in concentrations measured in billionths of a gram; yet, they are the principal effectors of critical cellular functions, such as cell proliferation, matrix synthesis, and tissue differentiation. Cytokines are similar to growth factors because they are receptor-activating locally acting polypeptides. Although cytokines were originally characterized from cells of the hematopoietic and immune cell systems, the distinction is rather artificial because most authors currently consider growth factors to be a subset of cytokines.
Although the same growth factor is often found throughout the body, they are named based on their function and tissue of origin. Several growth-promoting substances are identified in bone matrix and at the site of fracture healing. These growth factors are believed to play a role in the healing process. Among these are the transforming growth factor beta (TGF-), BMP, fibroblast growth factors (FGF), IGFs, and platelet-derived growth factor (PDGF). These growth factors are produced by osteoblasts and incorporated into the extracellular matrix during bone formation. Small amounts of growth factors can also be trapped systemically from serum and incorporated into the matrix. The present hypothesis is that growth factors are located within the matrix until remodeling or trauma causes solubilization and release of the proteins.
TRANSFORMING GROWTH FACTOR BETA
TGF-s are a family of dimeric polypeptide growth factors and coded by closely related genes. The family includes at least five molecules known as TGF- 1–4, and, by itself, it is a member of a superfamily that includes BMPs, activins, and GDFs, among others, that regulate morphogenesis in early development. The broad range of cellular activities regulated by TGF-s include the proliferation and expression of differentiated phenotypes of many of the cell populations that make up the skeleton. Among these are the mesenchymal precursor cells for chondrocytes, osteoblasts, and osteoclasts. The presence of TGF-s in normal fracture healing suggests they play a role in the repair process. TGF-s are secreted in an inactive form requiring acid pH or heat for activation. These dimeric, disulfide bonded molecules mediate their function through receptors that act as serine/threonine kinases. The receptors autophosphorylate, after forming a complex with TGF-, activate Smad intracellular pathways that translocate to the nucleus and regulate gene transcription. Thus far, eight mammalian Smad proteins are identified. After oligomerization, these proteins enter the nucleus to regulate transcription following assembly with transcriptional cofactors and comodulators. Because articular cartilage has limited potential for repair, TGF- 1 is used in animal studies for its stimulatory effect on chondrogenesis in periosteal explants.
BONE MORPHOGENETIC PROTEIN
Related to the TGF-s, the bone morphogenic proteins constitute a family of at least 15 growth factors originally identified for their ability to stimulate de novo bone formation. The BMPs are also referred to as osteogenic proteins (OPs); hence, there is some name confusion because BMP-7 is also called OP-1. BMPs are the only growth factors that can stimulate differentiation of mesenchymal stem cells into a chondroblastic (BMP-2) and osteoblastic (BMP-5, -6, and -7) direction. These BMP molecules are also quite effective osteoinductive agents. Noggin and chordin are extracellular binding proteins that alter binding of BMP molecules with their receptors. When implanted, most BMPs can stimulate a cascade of cellular events that closely mimics the process of endochondral ossification in normal fracture healing. Precursor cells are recruited and differentiated into chondrocytes that manufacture cartilage matrix. The cartilage is then gradually replaced by bone as osteoblasts populate the site. Eventually, bone marrow elements fill the newly formed intertrabecular spaces, and the bone remodels. Overexpression of BMP-4 in inflammatory cells is responsible for fibrodysplasia ossificans progressiva.
FIBROBLAST GROWTH FACTORS
Fibroblast growth factors are currently a group of eleven polypeptides that were originally discovered on the basis of their mitogenic effect on fibroblasts and have four known receptors. Acid fibroblast growth factor and basic fibroblast growth factor, also termed FGF-1 and FGF-2, respectively, are both implicated in cartilage and bone regulation. Basic FGF is generally more potent than acid FGF. FGFs have a significant proliferative effect on osteoblasts but less effect on protein synthesis. FGFs probably enhance bone formation by increasing the number of cells capable of synthesizing bone collagen. FGFs are also angiogenic factors, which are important for neovascularization during bone healing. A defect in the FGF receptor 3 is implicated as the cause for achondroplasia.
INSULIN-LIKE GROWTH FACTOR
The pivotal role of IGF in regulating endochondral ossification in skeletal growth suggests that this factor may also participate in endochondral ossification of bone healing. IGF regulates both bone matrix formation and cell replication. Of all the growth factors present in bone matrix, IGF-2 has the highest concentration; however, IGF-1 is four to seven times more potent than IGF-2. IGF-1, also known as somatomedin C, is produced in the liver and by skeletal tissue in response to stimulation with growth hormone. The cells that secrete IGFs also secrete any one of the six IGF-binding proteins that actually regulate the effectiveness of this growth factor.
Both IGF-1 and IGF-2 stimulate preosteoblastic cell replication by increasing the number of cells capable of synthesizing bone matrix. However, their mitogenic effect is less pronounced than those of other growth factors. IGFs also have independent effects on the differentiation of osteoblasts, increasing bone collagen production and inhibiting collagen degradation. Recent studies evaluating the endogenous tissue levels of different growth factors within healing tendon lesions show that there are low endogenous levels of IGF-I. As such, the exogenous administration of IG-1 during the first 2 weeks following injury may provide a therapeutic advantage by bolstering low endogenous tissue levels and thereby enhancing the metabolic response of individual tendon fibroblasts.
PLATELET-DERIVED GROWTH FACTOR
PDGF was discovered in serum as having a major mitogenic activity responsible for growth of cultured mesenchymal cells. It is a dimeric molecule that exists in two isoforms (A and B). When it is overproduced with some tumors, the protooncogene name is c-sis.PDGF is a potent regulator of bone cells and chondrocytes and plays a role in tendon healing.
VASCULAR ENDOTHELIAL GROWTH FACTOR
Vascular endothelial growth factor (VEGF), a potent angiogenic agent, is found in endothelial cells and responsible for the formation of new vasculature. It is involved with the angiogenesis of calcified cartilage and distraction angiogenesis. Although not significantly involved in the formation of normal bone, VEGF is involved with angiogenesis of malignancy. Clinical trials are being performed to evaluate the effectiveness of VEGF therapy into ischemic areas and the role of anti-VEGF therapy for treating malignancy. Studies show that administering autologous platelet-rich clots through the action of VEGF may be beneficial to the treatment of tendon injuries by inducing cell proliferation and promoting the synthesis of angiogenic factors during the healing process.
Extensive efforts are being made to find methods by which growth factors can be used to stimulate local bone healing and bone formation in a variety of clinical models. The growth factors TGF-, BMP, IGF-1, and basic FGF are currently the only growth factors that are demonstrated to possess substantial capacity for in vivo bone stimulation. Growth factors are probably best able to exert a stimulatory effect when used in conditions associated with impaired healing. Research has begun to show increasing evidence that growth factors can be used in vivo to stimulate bone healing and bone formation. The growth factors BMP-2 and BMP-7, also known as osteogenic protein-1 (OP-1), are in the final stages of pivotal human trials. Issues of the best delivery system are currently being debated from the use of collagen gel/sponge carrier to ex vivo adenoviral-mediated delivery systems.
There are many challenges to the clinical application of growth factors. It is unlikely that cell-signaling molecules act independently of one another or are present in isolation from one another at their sites of action. Although therapeutic measures that employ single agents may be efficacious in some circumstances, it is likely that most clinical applications will require the development of combination or serial treatment regimens. In addition, because the specific actions of growth factors are context dependent, it is critical to distinguish appropriate from inappropriate indications. Considerable effort is directed at altering the method of treating anterior cruciate ligament (ACL) ruptures. Certain growth factors such as TGF- 1, PDGF-AB, and FGF-2 can alter the biologic functions of human ACL cells in a collagen-glycosaminoglycan (CG) scaffold implanted as a bridge at the site of an ACL rupture. The addition of these selected growth factors to an implantable CG scaffold may actually facilitate ligament healing in the gap between the ruptured ends of the human ACL.
The therapeutic application of growth factors must also accommodate the fact that most factors have a widespread and variable distribution of target cells. A growth factor administered to elicit a desired response from one cell type may also influence other cell types, possibly in unintended or undesirable ways. Finally, in addition to demonstrating acceptable safety profiles and providing a physician-friendly delivery system in the current era of cost consciousness in health care, a growth factor treatment must demonstrate cost effectiveness along with clinical efficacy. Growth factors have many potential orthopedic applications. As challenges are met, it is plausible, if not probable, that growth factors will provide a means of treating patients with a variety of musculoskeletal disorders.
IMPLANT DESIGN & BIOLOGIC ATTACHMENT PROPERTIES
Total joint arthroplasty requires the type of implant materials and design that can support large functional loads. The implant must remain stable and rigid with respect to the bone while sustaining these loads. Adequate interface fixation requires interface micromotion of less than 100 m or gap spaces less than a fraction of a millimeter. Precise and uniform contact between the device and surrounding bone depends on the skill of the surgeon and the design of the instrumentation with which the site is prepared. The surface area of actual contact is probably small relative to the surface area of the implant. The contact points tend to induce areas of stress concentration rather than distribute the stress evenly. Bone maintains its structural integrity by responding to stress. In areas of stress concentration, bone resorption often occurs. Additionally, fibrous encapsulations of varying thicknesses are commonly found, and these further alter the ability of the implant to distribute stress uniformly. Implant loosening and migration may eventually occur and cause discomfort to the patient, in which case the implant may need to be removed.
Implant Fixation Mechanisms
Several types of implant fixation methods and surface texture designs were investigated to obtain better surgical fit and stress distribution at the implant-bone interface. The methods include the use of a grouting agent, direct bone apposition to the implant surface, bone growth into porous-surfaced implants, and chemical bonding between bone and surface-active ceramic implant coatings.
PMMA bone cement provides a mechanical interlock between the metal prosthesis and adjacent bone. Bone cement is not an adhesive; therefore, mechanical interlocking depends on the amount of interdigitation of the cement with trabecular bone and the quality of the fixation between the cement and the metal. The most favorable effect of cement fixation is immediate stability of the implant. Bone cement allows for load distribution over a larger area of bone, which reduces stress concentrations that may result in pressure necrosis and remodeling. Despite the early clinical advantages of cement fixation, the long-term results (15–25 years) are not as encouraging. The poor fatigue properties of cement lead to fractures of the cement. These fractures result in altered stress patterns in the bone, eventually leading to bone remodeling and implant loosening. Particulate debris that results from cement fracture is associated with osteolysis and aseptic loosening.
Improvements in the mechanical properties of cement and cementing techniques over the past two to three decades are leading to more promising results regarding the longevity of cemented prostheses. Femoral stems that were implanted with modern cement techniques maintain stability and function beyond 10 years in 95–98% of the cases.
DIRECT BONE APPOSITION
Optimal osseointegration at the bone-implant interface is affected by the material properties and design of the implant. Implant design encompasses both the surface texture and geometry of the implant. The mechanical properties of the implant-bone interface were investigated with various surface preparations, including smooth finishes, roughened or grit-blasted finishes, and grooved surfaces. Histologically, implants with smooth finishes have interfaces characterized by fibrous encapsulation, whereas implants with grit-blasted finishes have interfaces characterized by areas of direct bone apposition. Numerous studies demonstrated that surface texture is a significant factor in obtaining adequate implant fixation with direct bone apposition methods.
Evaluation of implant materials, including PMMA, commercially pure titanium, aluminum oxide, and low-temperature isotropic pyrolytic carbon with various surface finishes, demonstrated that the implant elastic modulus or surface composition did not significantly affect the interface attachment strength or histologic response. Surface texture significantly affected the interface mechanical properties, however. Implants with grit-blasted surfaces exhibited significantly higher interface attachment strengths than implants with polished surfaces did. Histologically, all implants with grit-blasted surfaces demonstrated areas of direct bone apposition, whereas all implants with polished surfaces demonstrated fibrous encapsulation (Figure 1–23), indicating that bone apposition required textured interface surface for attachment.
POROUS INGROWTH ATTACHMENT
The long-term problems associated with implant fixation with PMMA led to the development the porous coating as a method for permanent biologic fixation of prostheses. It is generally accepted that an implant can achieve stabilization by tissue growth into the surface porous structure if (1) the material is bioinert, (2) there is direct apposition of the bone at the implant interface, (3) there is minimal or no movement at the implant site, and (4) the porous structure has appropriate pore size and morphology.
Porous coatings are effective as a means of biologic fixation because their interface attachment strength of fixation is at least an order of magnitude higher than that of nonporous implants relying on direct bone apposition for fixation.
To maintain optimal bone growth into a porous structure, the pores must be large enough to accommodate the development of bone tissue. Several groups of investigators concluded that a pore size of 100 m allowed bone ingrowth but that a pore size greater than 150 m was necessary for osteon formation. Another group investigated the optimal pore size range for cobalt-based alloys by observing the rate of bone ingrowth and time necessary to attain maximal attachment strength. The results indicated that although a pore size range of 50–400 m obtained the maximal attachment strength in the shortest time, osteon formation was not demonstrated histologically at this range. In addition to a minimal pore size, an effective porous coating must also have appropriate pore morphology. The available porous layer for bone growth must be large enough to accommodate a sufficient quantity of bone to maintain adequate fixation, and a volume fraction porosity of 35–40% is accepted as optimal for effective biologic fixation of an implant with a strongly bonded porous layer. The volume fraction porosity is related to the interconnection pore size, particle interconnectivity, and particle size of the porous coating. Particle interconnectivity is important for ensuring adequate strength within the coating and between the coating and substrate. Too much particle interconnectivity can decrease the interconnection pore size and restrict the amount and type of ingrown tissue, however. A two-layer porous surface creates an interconnected and open porosity that is effective in creating a three-dimensional mechanical interlock of the ingrown bone.
Several different types of porous coatings were evaluated, including the fiber mesh, beaded porous, and irregular plasma-sprayed types. The fiber mesh type of porous coating is composed of wires that are cut and kinked to form the specific shape of the coating. The wires are then bonded to a solid metal substrate of the same metallic alloy through a sintering process in an inert gas environment or vacuum. The porosity obtained with this technique ranges from 40% to 50% with a mean pore size of 270 m. A void type of porous coating was obtained using a cobalt-based or titanium-based alloy. Magnesium microspheres are mixed with the base alloy by means of an investment casting technique. Under high temperatures, the magnesium evaporates, leaving pores on the surface of the alloy. This technique produces pores with different depths and connectivities. The most frequently studied porous coating is the beaded type. Cobalt-based alloy or titanium metal powder or macrobeads are either gravity compacted or applied with an organic binder onto a substrate. The beads are then sintered to the substrate at a submelting temperature for the base metal. The porosity ranges from 30% to 45% with pore diameters ranging from 100 to 400 m. Production of another type of porous coating involves plasma-spraying titanium to either a titanium-based or cobalt-based alloy substrate. The plasma-spray technique is further discussed with regard to ceramic coatings (see following section).
The extent to which implants are covered with porous coating varies. On femoral stems, it ranges anywhere from complete coverage to coverage of the proximal third. The extent of porous coating to achieve optimal stability is not determined. Circumferential coverage is preferable, however, to prevent wear debris migration toward the distal portion of the prosthetic stem.
Clinically, the short-term results for porous-coated prostheses are comparable to those for cemented prostheses. Histologically, retrieved human prostheses demonstrated variable amounts of bone ingrowth, ranging from limited to extensive, with large amounts of fibrous tissue (Figure 1–24). Some retrieved prostheses exhibited complete fibrous tissue infiltration into the porous surface. Components with limited bone ingrowth showed fibrous tissue that was oriented in a fashion capable of load transmission. The bone ingrowth and extensive fibrous ingrowth are most likely effective mechanisms for early implant stabilization. A histologic analysis performed on six retrieved, noncemented, porous-coated femoral hip components demonstrated a significant increase in bone ingrowth from 19 to 53 months after implantation. The data showed that human bone remodels slowly and advances appositionally with limited endochondral ossification. Therefore, to achieve reproducible bone ingrowth, the porous coating must be adjacent to cortical bone.
SURFACE TOPOGRAPHY AND COATINGS
Calcium phosphate coatings on metal surfaces were developed to overcome the mechanical problems of the ceramic as well as the biologic shortcomings of the metal. Many composite implants are manufactured to have the mechanical properties of the metal and the biologic properties of the bioactive ceramics. However, evidence is mounting from animal studies suggesting that quantity of ingrowth is approximately 80% dependent on the microscopic surface morphology and only modestly (approximately 20%) dependent on the hydroxyapatite (HA) chemistry. The bone formation is apparently encouraged by a microscopically rough surface, which is routinely produced with a crystalline HA surface coating. Titanium ingrowth was enhanced by acid etching the surface of implants, which changes its microscopic topography. Biological coatings of HA (calcium phosphates) may still enhance the chemistry of the surface, and they are the standard method of inducing ingrowth.
The bond between a metal substrate and a HA coating is critical to the success of a coated prosthesis. The processes and techniques of coating a prosthesis vary, and for this reason not all HA coatings perform equivalently.
A coating thickness of 50 m is generally accepted as adequate for coverage of nonporous surfaces; thinner coatings are necessary for porous coatings to keep from blocking the pores. Coating thickness >50 m tend to separate from the substrate and can cause debris that compromises the bearing surfaces. A coating thickness of 25–45 m does not occlude the pores or alter the mechanical bone ingrowth properties of the porous surface.
Chemical dissolution of an HA surface within the first few months of implantation occurs when coating thicknesses of 15 m or less are used. Calcium phosphate coatings are applied to substrates by a variety of methods, including dip coating, vacuum deposition, and plasma spraying. In dip coating, the substrate can either be dipped into a suspension of ceramic powder in a carrier or be dipped into a liquid form of glass ceramic. In vacuum deposition, ceramic material is removed from a source and deposited onto the target substrate.
Plasma-spraying methods are used by most manufacturers to apply calcium phosphate coatings to metal surfaces, particularly for load-bearing applications. A plasma or ionized gas is created by passing a gas or mixture of gases (usually argon or a mixture of nitrogen and hydrogen) through a high-energy, direct current (DC) electric arc struck between two electrodes. Then the coating powder suspended in a carrier gas is introduced into this plasma stream, melted, and propelled onto the substrate target, usually at high velocities. The coating is applied in several layers, each approximately 5–10 m thick.
Because of the high temperatures necessary for plasma spraying, the ceramic coating material may be chemically or structurally altered from the original ceramic. For this reason, all HA coatings are not identical and may vary among manufacturers in their composition, crystallinity, density, purity, and structure. These differences may affect the surface topography, bioactivity, and bioresorbability of coatings and make it nearly impossible to predict their long-term in vivo behavior. To ensure the correct composition of HA coatings, manufacturers perform a variety of tests, including radiograph diffraction, infrared spectroscopy, scanning electron microscopy, and atomic absorption spectroscopy. Several studies showed that HA-coated implants are superior to uncoated implants in terms of interface attachment strength and bone apposition. In newer research, HA coating improved bone purchase and bone-screw interface strength in healthy and osteopenic animals. Also, the properties of HA-coated implants with larger surface areas (macrotextured or porous surfaces) are superior to the properties of HA-coated implants with smaller surface areas (smooth surfaces). IGF-1, basic fibroblast growth factor (bFGF) and TGF- 1, alone and in combination, were tested to augment osseointegration and were absorbed onto a carrier of -tricalcium phosphate (-TCP). These composites were implanted into a defect around a hydroxyapatite-coated, stainless steel implant in the proximal tibia of rat in a model of revision arthroplasty. Although no growth factor combination significantly enhanced new bone formation or the mechanical strength of the implant, of the growth factors tested, only bFGF had any beneficial effect on the host response to the implant, perhaps by delaying osteoblast differentiation and thereby prolonging osteoclast access to the ceramic.
Clinical experience with HA-coated orthopedic prostheses is limited. Therefore, only short-term studies based on clinical follow-ups and radiographic evaluations exist. In a prospective bilateral total hip replacement study, a titanium prosthesis with and without HA coating was evaluated. At a mean of 6.6 years after implantation, there was no difference in clinical or radiographic results. Similar short-term results were reported for HA-coated, porous, cobalt-chromium primary and revision femoral hip stems. Other findings in early clinical studies of HA-coated porous implants included decreased pain, decreased radiolucent lines around the implant (Figure 1-25), and improved bone remodeling.
Not all studies in patients with HA-coated implants have reported positive findings. In fact, some newer studies reported cell-mediated osteolysis, implant loosening, and other negative effects linked with the degradation or delamination of the HA coating, the generation and migration of HA particles, and the subsequent three-body wear of the implant that is caused by these particles. These findings suggest that changing the surface topology might be an effective way of achieving the same or nearly the same goal with less potential problems.
Factors That Affect Biologic Attachment
Attachment at the bone-implant interface is affected by the material properties and design of the implant, surgical technique, initial implant stability, and direct contact with the surrounding bone. Initial implant stability and apposition with bone are not always achievable but are vital for implant longevity. Persistent micromotion at the bone-implant interface causes bone resorption and necrosis, which can in turn result in fibrous tissue infiltration at the interface and in implant loosening. Moreover, any initial gap between the implant and surrounding bone may adversely alter the amount of osseointegration and the rate at which it occurs.
MOTION AT THE BONE-IMPLANT INTERFACE
Motion of an implant within the surgical site has a primary influence on biologic fixation and implant longevity. Initial implant stability is essential for the early tissue infiltrate within the porous structure to differentiate into bone by either direct bone formation or appositional bone growth. When excessive early movement occurs at the bone-implant interface, bone formation within the pores is inhibited. The majority of research concerning motion at the interface involves porous implants; however, the findings are applicable for press-fit implant systems.
Studies of implants suggest that relative motion of greater than 150 m at the bone-implant interface prevents bone formation, although a well-ordered fibrous tissue interface was maintained and provided adequate implant attachment. When interface motion was 40 m, bone ingrowth occurred, but the calcified ingrown bone was not continuous with the surrounding bone. These findings support the concept that in optimal bone growth into porous surfaces or bone apposition onto press-fit surfaces requires little or no initial micromotion at the interface.
In a study of HA coating in a continuous loaded implant model, investigators found that when interface motion led to the formation of a fibrous membrane, the HA coating was able to convert the membrane to bone. Furthermore, pulsed electromagnetic fields accelerated HA osteointegration in trabecular bone.
The technical difficulties in cutting bone precisely to provide an exact fit around the implant often result in a poor surgical fit. Implant and instrumentation design may also make it difficult to achieve initial implant-bone interface apposition. When a femoral stem is press-fit into the femoral canal, only 10–20% of the prosthesis comes into direct contact with bone.
The effects of interface gaps and poor surgical fit of implants were investigated by numerous groups. In studies of HA-coated prostheses, researchers found that the HA coating does not compensate for improper implant placement or poor surgical technique. The cell populations necessary for bone formation are identical across large interface gaps and in press-fit situations. In large gaps, the rate of gap filling and subsequent ingrowth is delayed, and the quality of bone at the interface may also be reduced. Such studies indicate the potential benefits of robot-controlled surgery, where much better initial apposition can be obtained.
TISSUE RESPONSE TO IMPLANT MATERIALS
The effect of an implanted material on adjacent tissues depends on the amount and type of substance released into the tissues, the histologic response to the material, and the wear and corrosion properties of the material. The type of response to the implant determines the biologic classification of the material.
Biotolerant materials, such as stainless steel, PMMA, and UHMWPE, elicit the worst tissue response. When these materials are used, a fibrous tissue layer may form between the bone and the implant. This fibrous layer is generally observable as a radiolucent line on radiographs. Examination with light microscopy shows the presence of numerous macrophages near resorbing adjacent bone and the resulting fibrous tissue that contains macrophages and foreign body giant cells.
PMMA elicits adverse local and systemic effects from the moment of its introduction into the body. At the time of implantation, PMMA causes local tissue necrosis because of the extreme heat of polymerization. During the polymerization process, monomer may leach into the surrounding tissues and cause hypotension. Finally, PMMA fragmentation particles elicit a chronic macrophage response at the implant-bone interface, which can result in progressive osteoclasis and eventual aseptic loosening of the implant. Macrophages are stimulated by cell necrosis, bacteria, and foreign particulate matter. Particulate matter is the primary cause of aseptic loosening in cemented joint arthroplasties. In spite of this, bulk PMMA is well tolerated by the body, whereas particulate PMMA is not.
Bioinert materials, such as titanium and cobalt-chromium alloys, usually cause minimal tissue irritation. With stable implants of either titanium or cobalt-chromium alloys, appositional bone growth or osseointegration occurs. If titanium implants are used in articulating surface applications, however, they have poor wear resistance, and the excessive wear particles behave as biotolerant materials. These particles elicit a chronic macrophage response, which can lead to implant loosening. However, newer animal studies showed that coating titanium implants with autologous osteoblasts accelerates and enhances the osseointegration of these implants and could be a successful biotechnology for future clinical applications.
Bioactive materials, such as calcium phosphate ceramics, offer the best biologic advantage of implant materials. The biocompatibility of the calcium phosphate ceramics is well documented. In response to these implanted ceramics, the body typically responds (1) without local or systemic toxicity, (2) without inflammatory or foreign body reaction, (3) without alteration of natural mineralization processes, (4) with functional integration of bone, and (5) with chemical bonding to bone via natural bone cementing mechanisms. Implant surfaces coated with HA are characterized as being capable of forming direct, intimate bonds with the surrounding bone. The bonding area (approximately 50–200 m) contains biologic apatite crystals that are highly oriented at the interface with a 10-m periodicity similar to that of calcified tissue, as determined by electron diffraction studies. The bone apatite crystals are arranged against the implant surface in a palisade fashion, resembling the natural bonding between two bone fragments. The bonding area contains a ground substance that is heavily mineralized, although devoid of collagen fibrils, and is likened to the natural bone cementing substance, which is amorphous in structure, heavily mineralized, and rich in mucopolysaccharides.
One study used animal models to characterize tissue-specific reactions to particles of bone-substitute materials for osteocompatibility. Tested particles included demineralized bone powder (DBP), nonresorbable calcium phosphate (nrCP), PMMA, polyethylene (PE), and resorbable calcium phosphates (rCPs). Although both DBP and nrCP were incorporated into the reactive medullary and cortical bone, DBP also induced enchondral osteogenesis, and nrCP evoked a fibrous reaction. Although PMMA particles were surrounded with a fibrous layer, they did not impair bone healing. PE shards and rCPs were inflammatory and inhibited osseous repair. rCPs, PMMA, and PE shards all generated inflammatory reactions with each particle being surrounded by fibrous tissue and large multinucleated giant cells. DBP showed both osteoinductive as well as osteocompatible properties. Although nrCP was shown to be osteocompatible, rCPs stimulated various degrees of inflammatory responses. PMMA was osteocompatible and did not interfere with the bone healing process. PE was not osteocompatible and generated foreign body reactions in both sites. It is important to distinguish among the osteoinductive, osteocompatible, and inflammatory properties of particles that may be used as bone-substitute materials.
Problems Associated with Maintaining Implant Longevity
Implant loosening, which can result from bone loss that is caused either by stress shielding or by osteolysis, has been a problem associated with total joint arthroplasty since its inception. Periprosthetic osteolysis presents radiographically as diffuse femoral cortical thinning or as a focal cystic lesion.
Although the exact cause of osteolysis is unknown, it is thought to be a result of movement of the implant, primarily at the bearing surface, with generation of wear particles that migrate to the implant-bone interface, where they cause a tissue reaction. Movement at taper joints, or between the implant and bone cement, can cause particulate material. Particulate debris in the 0.1–1.0 m range are most active in stimulating bone resorption and eventually causing implant loosening. Macrophages play a key role in the osteoclastic process by elaborating cytokines in host defense and in forming precursors for osteoclasts. TNF- is a key cytokine in the osteolysis process. The prevalence of osteolysis in stable cemented femoral components ranges from 3% to 8%. It appears, however, that osteolysis is observed earlier in patients with stable uncemented components and that the prevalence increases with time in vivo. The prevalence in uncemented systems ranges from 10% to 20% after 2–9 years in vivo. Newer studies suggest that hydrostatic pressure may play a far greater role in the induction of osteolysis than previously thought.
In addition to osteolysis, another problem with total joint implants is the increase in metal ions released into the body. This problem is especially associated with uncemented porous-coated implants. Systemic and long-term effects caused by wear and corrosion are just being discovered. An understanding of the wear and corrosion mechanisms associated with decreasing implant longevity is vital for the development of improved implant designs and material manufacturing methods.
SURFACE DAMAGE OF POLYETHYLENE IMPLANTS
Osteolysis, loosening, and other complications that reduce implant longevity are attributed to polyethylene wear particles. Careful examination of retrieved polyethylene components demonstrated a variety of modes by which surface damage occurs. These include scratching, burnishing, embedding of debris, pitting (the presence of shallow, irregular surface voids in the surface), surface deformation (permanent deformation on the articulating surface), abrasion (characterized by a tufted or shredded appearance of the polyethylene), and delamination (separation of large, thin surface sheets of polyethylene from implant components).
Fatigue is suggested as the primary mechanism of polyethylene surface damage because the damage was correlated with the length of time since implantation (number of cycles in the fatigue curve shown in Figure 1–2) and with patient weight (applied load or stress in Figure 1–2). Surface damage is noticeably less in acetabular components than in tibial components. The increased polyethylene damage in total knee arthroplasties can be attributed to reduced surface conformity and to compression-tension loading patterns. Nonconformity of articulating components in total knee arthroplasties results in contact stresses that approximate or exceed the yield strength of the polyethylene. Cruciate-retaining designs vary the location of contact over the entire articulating surface, thereby subjecting the implant components to alternating compression-tension contact stresses throughout the loading cycle. This cyclic process could contribute to the beginning and spread of cracks, which may lead to pitting, delamination, and other fatigue failure modes.
The elastic modulus and thickness of the polyethylene are significant predictors of contact stresses large enough to cause surface damage. An increased elastic modulus, as is found with enhanced (Hylamer) polyethylene, raises contact stresses and can be expected to result in increased wear. This is important in component design because the elastic modulus of polyethylene near the surface may increase up to 100% over 10 years in vivo. A reduced level of polyethylene thickness, as is found in metal-backed acetabular and tibial components, can result in increased wear and creep, eventually leading to cracking and separation of the polyethylene from the metal. To avoid the high stresses that cause cracking, acetabular polyethylene thickness should be greater than 6 mm, and tibial polyethylene thickness should be greater than 8 mm. Another concern regarding metal-backed components is that loosening of the metal backing may cause the screws to break and migrate into the polyethylene insert, and this in turn can result in the generation of large amounts of metal and polyethylene wear debris.
The current practice is to try to minimize polyethylene wear debris through the use of highly cross-linked polyethylene for total hip applications, with some trials for total knee applications. The higher contact stresses on knee components is concerning from a materials viewpoint because the cross-linked material has a lower ductility and fatigue strength, although even in this application, early reports are encouraging. Other strategies to reduce particulate wear include alternative bearings, such as ceramic on ceramic, metal on metal, and ceramic on polyethylene, either as bulk ceramic (alumina) or as a surface treatment (oxidized zirconium).
FATIGUE OF POROUS-COATED IMPLANTS
The primary failure mode of load-bearing orthopedic implants is fatigue. The majority of hip and knee systems have sintered porous coatings to maximize biologic fixation or cement impregnation. The fatigue properties of these porous-coated implants are influenced not only by the sintering treatment but also by a notch effect from the coating.
Sintering affects the fatigue properties of coated implants by altering the microstructure of their metal substrate. With titanium-based alloys, sintering requires that the material be heat treated above the beta phase transition temperature, which reduces the fatigue properties of the material by approximately 40%. When postsintering heat treatments are performed, the fatigue strength of the previously sintered titanium-based alloy increases by 25%. With cobalt-based alloys, sintering does not necessarily result in a reduction of fatigue strength. A dissolution of carbides and an increase in porosity occur, however, when cobalt-based alloys with less than 0.3% carbon are exposed to sintering temperatures. Additionally, with improper cooling, the sintered cobalt-based alloys can develop continuous grain boundary precipitates.
Investigators performed studies to determine the effects of sintering, postsintering heat treatments, and HIP techniques on the fatigue properties of nonporous and porous-coated cobalt-based alloys. They reported that sintered materials exhibited severe porosity and continuous grain boundary precipitates, which resulted in reduced fatigue and tensile strength. HIP eliminated the porosity and grain boundary precipitation resulting from sintering, however. Moreover, HIP of the sintered materials increased the tensile and fatigue properties in implants with or without a porous coating.
Aside from manufacturing processes, which may alter the fatigue properties of the substrate, porous coatings demonstrate a notch effect at the contact regions. These regions are susceptible to the initiation of cracks, which may continue to propagate along surface grain boundaries. This effect is most significant for the titanium-based alloys.
The fatigue properties that are caused by sintering and the notch effect of coating in load-bearing implants can be reduced by the following measures: (1) avoiding the use of porous coating in regions of maximal tensile stress, (2) using an additional heat treatment process on previously sintered titanium-based alloys, and (3) using HIP on previously sintered cobalt-based alloys.
The fatigue properties have to be considered in relation to newer animal studies that evaluated the roles of sex and estrogen therapy on the amount of bone ingrowth into porous cobalt-chromium implants. Histological examination showed significantly more bone ingrowth in areas with cortical bone contact than in areas with cancellous bone contact, with no difference between male and female animals. However, ovariectomized animals showed less overall bone ingrowth than male and female controls, and bone ingrowth in areas with cortical bone contact did not decrease significantly, whereas bone ingrowth in areas with cancellous bone contact was significantly impaired. In another study evaluating osseointegration, osteoblasts from patients older than 60 years were less able to form bone on Ti-6Al-4V implants. Taken together, these data suggest that extensively coated or full-coated porous prostheses are recommended to achieve enough cortical bone contact and ingrowth for postmenopausal patients.
METAL IMPLANTS: ION RELEASE, METAL-ON-METAL BEARINGS, AND IMMUNOLOGY
Any metal exposed to the physiologic environment will corrode. Corrosion is most evident in fracture fixation devices. Retrieval studies of stainless steel components revealed evidence of pitting and crevice corrosion in approximately 75% of the components.
Apart from potential implant mechanical failure, the clinical significance of corrosion is determined by the type and quantity of metal ion that is released and by the local and systemic effects of ion release. The widespread use of porous coatings on metallic implants and the popularity of metal-on-metal bearing surfaces has raised additional concerns regarding metal ion release. Porous coatings increase the amount of surface area that is exposed to body fluids by a factor of 1.2 to 7.2. Depending on the type and morphology of the porous coating, the increased surface area could increase the corrosion and ion release rates by a factor of 1.2 to 5.2. Unlike the metallic surface of cemented implants, the metallic surface of porous implants is in direct contact with the endosteal bone surface and vasculature, creating an environment in which a cellular response to metallic ions is possible. Another significant source of metal ions in the systemic circulation is the loose prosthesis. Local synovial fluid levels were increased in most prostheses at revision, but examination of blood showed that metal ion concentrations were elevated only in cases involving loose components. The metal-on-metal (MOM) bearing does not directly increase the corrosion area but does produce metal particulate debris that has a very large surface area to volume ratio. This debris then is available to dissolve and put ions into the systemic circulation. Significantly increased levels of implant elements are documented in serum and urine after implantation of these devices. Reference values for two of these elements in blood are cobalt, <0.15 g/L, and chromium, <0.26 g/L. Blood cobalt levels are reported to be 5–20 times higher and chromium levels 5–10 times higher than reference levels for MOM bearing total hips.
Released metal ions (Al, Co, Cr, Fe, Mn, Ni, Ti, and V ions) have one of four types of possible effects on the body: metabolic, bacteriologic, immunogenic, or oncogenic. With the possible exception of titanium, the metallic elements used to fabricate implants are either essential or toxic to processes of metabolism. Although it is known that excessive concentrations of essential elements can produce toxic effects, the ultimate fate of the released ions locally, systemically, or in remote organ systems remains to be determined. There is a reported increase in chromosome translocations and aneuploidy in peripheral lymphocytes after MOM hip arthroplasties, but the risk of cancer does not seem to be increased in the few underpowered studies available.
Metal sensitivity induced by metal ion release is not a common problem causing prosthesis loosening. The prevalence of dermal sensitivity in patients with joint replacements is greater than in the general population, and dermal patch testing is not accepted for identifying patients at risk for hypersensitivity to the implant materials. Nickel, chromium, and cobalt (present in cobalt-chromium [Co-Cr] alloy) are the ions most frequently responsible for metal sensitivity reactions. When investigators studied porous-coated hip and knee devices retrieved from patients, they reported the finding of a cellular response in interfacial and interstitial tissues. Although none of the components were removed because of infection and none had shown clinical or radiographic evidence of loosening, the investigators identified an inflammatory infiltrate with accompanying vascular proliferation in 22% of the components. The predominant cell types within the porous coatings were lymphocytes and histiocytes, although giant cells were also present. Several groups reported that delayed hypersensitivity can produce an immunologic reaction in which T cells recognize a metal ion-protein complex, release a variety of lymphokines, and stimulate a mononuclear infiltration. Further research is necessary, however, to determine if an allergic or hypersensitivity response to metal ions is responsible for inflammatory infiltrates. Various tests were proposed to determine if a particular patient is likely to have a sensitivity reaction. However, sensitivity is rare, so routine testing would be expensive and unrewarding, and in the case of a positive history, the surgeon would still be likely to use a component that does not contain the potentially offending elements, such as titanium alloy or zirconium alloy, with ceramic as a femoral head.
MOM bearing surfaces, on the one hand, have the potential to reduce the wear debris problem, but, on the other hand, they are certain to raise levels of circulating ions and raise issues of long-term safety from an organ viewpoint. Hypersensitivity issues may also be increased.
Gait Analysis: Introduction
Author: Harry B. Skinner, MD, PhD
The science of studying human walking is called gait analysis. This science evolved as a means of quantitating the individual components of gait. As measurement techniques were refined to permit the determination of forces, moments, and movements of the human body, these techniques were applied to functions other than walking. Thus, it has been possible to measure the functional demands of wheelchair motion and running, as well as activities as diverse as pitching a baseball. The study of gait analysis was assisted by the development of devices that are able to measure gait in terms of (1) movement in space, (2) metabolic energy consumed during movement, (3) functional patterns of muscles during movement, and (4) forces applied to the ground during movement. Direct measurements of these factors permit the secondary determination of quantities concerning mechanical work, joint moments, and center of pressure, which in turn are helpful in quantifying the function of prostheses and the effects of ataxia. The techniques of gait analysis are applied to other facets of human function. Motion analysis, for example, is applied to elucidate the kinesthetic changes in the spine occurring with fatigue in a study of repetitive lifting.
GAIT CYCLES, PHASES, & EVENTS
For uniformity in the reporting of gait measurements, investigators adopted several definitions concerning gait cycle. One gait cycleis defined as the time from initial ground contact of one foot to subsequent ground contact of the same foot. This is then normalized to 100%, with the intervening phases, periods, and events (Figure 1–26) defined on this basis. Ground contact is chosen as the beginning of the cycle because it is easily defined. The duration of the gait cycle varies, depending on the height, weight, and age of the individual whose gait is being analyzed as well as on any pathologic process affecting the individual's movement. Normalization of the gait cycle into percentages facilitates comparison among individuals.
Two gait phases are recognized: the stance phase and the swing phase. The normal stance phase, when the extremity is on the ground, accounts for approximately 62% of the cycle, and the normal swing phase of that extremity accounts for the remainder. The proportions vary with speed. Each phase is divided into periods. The stance phase starts with double-limb support, is followed by single-limb support, and ends with a second period of double-limb support. Each period of double-limb support accounts for an estimated 10% of the gait cycle. The swing phase is divided into early, mid, and late periods, with each period accounting for approximately 13% of the gait cycle. Swing phase for one limb corresponds to stance phase for the opposite limb.
The swing and stance phases are also divided into gait events (see Figure 1–26). Terms such as heel strike and toe-off, which are used to describe the events in the gait cycle, were initially derived from the observation of normal gait. The nomenclature of gait analysis, however, is evolving to take into account the fact that these terms are inadequate in describing the gait of individuals who have joint contractures, joint instability, pain, spasticity, or other conditions that alter the gait so the heel may never strike the floor or the heel and toe may strike simultaneously or depart simultaneously. In gait analysis, the various events are recognized by observation and can be correlated with measurements of the ground reaction force and motion variables. Observation is greatly enhanced through the use of slow-motion photography or video equipment.
The quantities measured in a complete analysis of gait include three-dimensional translation, velocity, acceleration for all motion segments, forces exerted on the ground, electromyographic response of muscles during the gait cycle, and metabolic energy consumption. Obviously, a complete gait analysis is an expensive and time-consuming procedure and perhaps not even possible within the endurance limits of some patients studied. A typical gait analysis is problem oriented and focuses on information relevant to the disorder being addressed by the clinician.
Movement during Gait
The fundamental data needed for almost any gait analysis are basic measurements termed stride characteristics. These are necessary because they form a baseline for interpreting all of the other aspects of gait. The stride characteristic variables are velocity (speed), gait cycle, cadence, stride length, step length, single- and double-limb support time, and swing and stance time.
Velocity of gait is the measure of forward progression of an individual's center of gravity, which is generally located midline and anterior to the sacrum. Velocity is expressed as an average number of meters per minute, although it is obvious that the instantaneous velocity can vary somewhat.
Gait cycle is measured as the number of seconds from the initial ground contact of one foot until the subsequent ground contact of the same foot. Cadence is the number of steps per minute (the number of times both feet strike the ground per minute) and is different from the number of strides per minute (the number of times the same foot strikes the ground per minute).
Step length is measured as the distance (number of meters) covered from the time one foot strikes the ground until the opposite foot strikes. It differs from stride length, which is the distance (number of meters) covered from the time one foot strikes the ground to the next time the same foot strikes the ground. In normal individuals, the length of each step is half of the stride length. But in people with pathologic processes that affect gait, the lengths of the two steps are different.
Single-limb and double-limb support times are periods of the gait cycle that can be measured in terms of seconds or in terms of percentage of the gait cycle (Table 1–4 and Figure 1–26). These periods in patients with a painful condition such as an ankle sprain generally differ from periods in normal individuals. Less obvious is the fact that the two double-limb support times (one following left foot strike and the other following right foot strike), which are usually of the same length in normal individuals, may be of two different lengths in patients with pathologic processes that affect gait. Swing and stance times can also be measured in terms of seconds or in terms of percentage of gait cycle.
Gait characteristics of normal men and women at free walking speed are shown in Table 1–4, and a sampling of these data for normal children at selected ages is presented in Table 1–5. Gait measurements are generally made at the free walking speed for each person because that speed is selected by the person to minimize energy consumption and is therefore considered the optimal gait velocity. Velocities that are slower or faster can be continuously maintained by individuals as long as the metabolic energy consumption remains in the aerobic range. These velocities are more costly in terms of energy expenditure, however.
Stride characteristics are sensitive indicators of diseases and disorders that affect gait. Many variables have a bearing on the stride measurements. Age, height, weight, and shoe wear are physiologic variables that help define the basic parameters of velocity, cadence, and gait cycle. Abnormalities resulting from an anatomic change, such as joint replacement, degenerative disease of the lower extremities, or knee fusion, can be demonstrated as nonspecific and asymmetric variations in the stride characteristics. External variables such as those affecting the walking surface (eg, sand, concrete, and ice) can markedly alter stride measurements and must be considered in comparing data from different treatment groups or locations. Data are also sensitive to measurement technique. "Free" walking behavior in a laboratory may be different from that which takes place unobserved on a street and is definitely different from walking on a treadmill. Thus, to eliminate extraneous variables, care must be exercised not only in measuring stride characteristics but also in interpreting them.
Motion analysis is necessary for the complete characterization of gait. Aside from simply recording motions, quantification of the dynamic range of motion of a joint or body segment is called kinematics. Rather than measuring all limb segments, investigators generally focus on motion in certain limb segments and the trunk. Although major displacements of the lower extremities and upper extremities are occurring during gait, the center of mass of the body is only moving approximately 2–4 cm in a mediolateral direction and 2 cm in a superoinferior direction. Simultaneously, pelvic and trunk motion is occurring around the center of mass in a sinusoidal fashion. To conserve angular momentum, the upper extremity moves forward with the contralateral lower extremity.
Motion analysis has benefits and limitations. On the one hand, it provides more information to the clinician than does simple analysis of stride characteristics. For example, although stride analysis may show that the single-limb support time is reduced, motion analysis can clarify whether this is caused by decreased hip or knee motion, weakness of knee or ankle musculature, or some other condition, and it can also permit documentation of the benefit of intervention. On the other hand, motion analysis equipment is expensive, although this is improving, as is the need for human input to the data analysis. It also presents difficulties in defining motion segments accurately and sometimes in measuring relatively small movements. There are problems, for example, in placing the markers to determine mediolateral pelvic rotation or anteroposterior pelvic rotation and in measuring these movements, whereas measuring knee flexion and extension is much easier. Highly sophisticated motion analysis systems were developed to maximize the accuracy and efficiency of these measurements, with a significant improvement. One study was able to discern abnormal tibial motion during gait in ACL-deficient knees.
Motion analysis has taken on new functions with the interest in upper extremity motion, especially in sports. Training problems and injuries in baseball pitching, golf, and racquet sports, among others, can be defined in terms of motions (rotations, displacements, and angular velocities) of the upper extremity joints. For example, world-class tennis players achieve ball velocities of 40–50 meters per second (m/s), with elbow extension and shoulder internal rotation angular velocity peaks of 1510 degrees per second (deg/s) and 2400 deg/s, respectively. Motion analysis can lead to methods of prevention and treatment, as well as improvement in technique.
Energy Consumption during Gait
Energy expenditure results from muscle function and is possible as a direct result of the body's use of food. Steady-state aerobic metabolism is the optimal means of using oxygen to metabolize food, although less efficient anaerobic mechanisms are available. Measurements of oxygen consumption per unit of time can be converted to measurements of energy expenditure or power. An oxygen consumption measurement of 1 L/min is approximately equivalent to an energy consumption measurement of 5 kcal/min.
Energy expenditure per unit of body mass can be expressed per step, per unit of distance, or per unit of time. Most commonly, energy expenditure is measured in terms of rate of oxygen uptake, expressed as milliliters of oxygen per kilogram per minute (mL O2/kg/min), and net oxygen cost, expressed as milliliters of oxygen per kilogram per meter (mL O2/kg/m). Both the rate of oxygen uptake and the net oxygen cost depend on the velocity (v) of walking, expressed in terms of m/min. The approximate relations are as follows:
The rate of oxygen uptake for normal adults is 11.9 ± 2.3 mL O2/kg/min, and the net oxygen cost is 0.15 mL O2/kg/m.
The rate of oxygen uptake increases with the square of the velocity (Figure 1–27). Body mass is obviously important in determining energetics, but location of the mass is even more important. An increase in weight around the center of mass (ie, the waist) is not nearly as energy costly as an increase around the ankles. This is because the center of mass moves at a near constant velocity with relatively small motions. Conversely, the ankles must be accelerated and decelerated constantly during gait, with each acceleration and deceleration requiring energy. When net oxygen cost for normal individuals is expressed as a function of velocity, there is a minimum in the energy consumption curve at approximately 80 m/min (1.3 m/s) at approximately 0.16 mL O2/kg/m (approximately 1.1 J/kg/m), indicating that this is the most efficient velocity of ambulation (Figure 1–28). Similarly, the most efficient running speed is between 3 and 4 m/s at an energy consumption rate of approximately 3.1 J/kg/m.
Energy expenditure in gait assumes importance when gait efficiency decreases or when the most efficient attainable gait velocity is markedly below normal. Attempts to increase velocity can increase energy costs to the point that sustained ambulation cannot be maintained. This can be seen, for example, in the case of traumatic transfemoral amputation. Although the amputee may ambulate with a net oxygen cost of approximately 0.28 mL O2/kg/m, which is nearly 100% above normal (0.15 ± 0.02 mL O2/kg/m), attempts to increase the speed could push the amputee into the anaerobic consumption range and thereby limit the ambulation distance. A similar problem can be seen in the case of a paraplegic who has adequate muscle function to ambulate before gaining weight but finds that the net oxygen cost of increased weight makes wheelchair mobility more energy efficient.
Muscle Function during Gait
Measurement of the function of muscles during gait is helpful in understanding and treating problems associated with cerebral palsy, stroke, poliomyelitis, and other diseases that alter the normal pattern of muscle function. The activity of muscles during gait is determined by dynamic electromyography through the use of either surface electrodes or fine-wire electrodes inserted into muscles. Electrical activity generated from these muscles is monitored and recorded in an on-off fashion as a function of the gait cycle. Activity does not necessarily indicate agonistic (contracting) or antagonistic (lengthening) function, but this determination can be made with simultaneous motion analysis. At present, it is not possible to quantify the relationship between electromyographic activity and force.
Normal functioning of the muscles can be presented as a function of the gait cycle and is shown in Figure 1–29. Most muscle activity is generated at the beginning and end of the stance and swing phases of gait because it is necessary at these times to accelerate and decelerate the extremities.
Dynamic electromyography is particularly useful in disorders associated with spasticity, such as cerebral palsy and cerebral vascular accident. In these disorders, the results of functional muscle testing of the patient in the supine position can be markedly different from the results of testing in the upright ambulating position. For example, the tibialis anterior may function normally when the patient is supine, but dynamic electromyography may reveal a varus-deforming force of the hindfoot during ambulation.
Forces during Gait: Kinetics
Forces acting during ambulation arise from gravity, inertia, and ground reaction. At ambulation speeds, viscous drag can be ignored. The gravitational force (mass x gravity) must be considered because it causes moments around centers of rotation for limb segments and body segments. The inertial force is proportional to the acceleration of the body segment and acts in the opposite direction because it resists acceleration. The ground reaction force is a measurement of the load applied to a device such as a force platform and has three components: vertical ground reaction force, fore-aft shear, and mediolateral shear.
Typical curves for the three components of the ground reaction force are shown in Figure 1–30. The dip in the vertical force curve (Figure 1–30A) during the single-limb stance phase of gait occurs because the inertial forces reduce the ground reaction force below body weight. The fore-aft shear (Figure 1–30B) is negative after the heel strike because the foot is pushing the plate anterior. On toe-off, the converse is occurring so the shear force is in the opposite direction. Again, correlation of ground reaction forces to stride characteristics can be beneficial in interpreting gait data. Force platform data show variations with walking speed, shoe wear, and compensatory mechanisms of gait, such as the avoidance of weight bearing on a painful extremity. The components of the ground reaction force in gait can be an indication of dynamic aspects of gait. The vertical ground reaction force tends to increase in magnitude and higher frequency content with increasing velocity and flatten with lower velocity and/or lower extremity pain. Examination of the frequency content especially in the medial-lateral shear component is an indicator of balance control, as in scoliosis. Static measurements provide center of pressure variations, also as a measure of balance control.
More sophisticated force measurement devices were developed to permit measurement of pressure (force per unit area). These devices can yield both in-shoe pressures and pressures applied to the external surface of the shoe by a force platform. Thus discrimination of measures at the 1-cm2 level is possible, allowing studies of feet of diabetic patients to discern differences shoe wear can have on the risk of skin breakdown.
ROLE OF GAIT ANALYSIS IN THE MANAGEMENT OF GAIT DISORDERS
Gait analysis was traditionally been a research tool and continues to find its primary use in the research arena. However, the impetus to document the benefit of medical intervention has resulted in a reliance on gait analysis to quantify gains made by surgery or other treatments. There continue to be proponents of gait analysis as a clinical diagnostic tool. New areas in which gait analysis is under investigation include Parkinson's disease, cervical myelopathy, high-heeled gait, and even depression. The publication of normative data permits the definition of pathologic gait and thereby defines a goal in the rehabilitation of a patient with a gait disorder. Analysis in gait laboratories can measure the initial deviation from normal as well as the improvements occurring through the rehabilitation process. Even as gait laboratories assume a more prominent role in evaluation of patients, however, the data generated from research laboratories have already affected clinical practice in a variety of ways.
New techniques of analysis were developed as computer sophistication and availability evolved. Gait is a process with inherent variability, which must be understood and accounted for to maximize its utility. Fractal analysis is one approach and may result in a better description of gait rhythm variations during maturation from childhood to adulthood. Understanding and predicting muscle activation patterns can be accomplished using artificial neural network models. These are particularly exciting because their use may lead to active control of prostheses. Part of the variability of gait arises from differences in body size. Scaling the gait data to body size can reduce intersubject variability and provide a tighter evaluation of abnormality. Identification of "abnormal" gait can be subtle in some cases in view of the variable and voluminous data generated. The common method of evaluation is principal component analysis, which quantifies the deviation of variables from normal. A newer report evaluates variability based on an index that used the squared distance from the mean obtained from joint angle measurements of many subjects. These techniques promise to improve the usefulness of gait analysis in the evaluation and treatment of orthopedic patients.
Gait studies of patients with transfemoral and transtibial amputations resulted in an objective evaluation of ambulatory potential after prosthetic replacement. Early studies demonstrated that amputations at more proximal levels resulted in greater loss of symmetry and increased energy expenditure (Table 1–6). These findings stimulated renewed attempts to maintain amputations at the most distal level. Other studies showed that strengthening the muscles of the residual limb in patients with transtibial amputations resulted in velocities that were improved although still lower than normal. Recent electromyographic studies of gait in transfemoral amputees demonstrated the importance of reattaching biarticular muscles that are transected at the amputation site to bone to maintain their function.
Techniques for gait analysis of amputees also provide the objective means of comparing various prosthetic components, such as the solid ankle cushion heel (SACH) foot, and the multiaxial and uniaxial feet.
Gait analysis is now a common tool in the treatment and evaluation of patients who have cerebral palsy or have suffered from a cerebrovascular accident or head injury. For disorders such as these, gait improvement can be expected for 6 months or more. Studies suggest that the natural history of gait in child and adolescent cerebral palsy is one of deterioration with aging, an important finding when evaluating long-term effects of intervention to improve gait. The use of electromyographic gait data can improve the results of muscle transfers in this group of patients. Knowledge of the activity period of a muscle (swing phase/stance phase) can allow prediction of the effect of transfer of that muscle. Thus, gait analysis is applied to spastic equinovarus feet to determine the efficacy of transfers of tibialis anterior and tibialis posterior. Similarly, gait analysis can be used to evaluate different surgical interventions, such as selective dorsal rhizotomies and orthopedic surgical interventions in spastic diplegic patients, showing improvement in energy cost and gait kinematics with both interventions.
Clinicians can use gait analysis to provide objective clinical data whenever orthotics are prescribed. Gait analysis before and after the application of an orthosis can quantify the effects in an objective manner. In evaluating the results, clinicians should remember that an orthosis that eliminates motion of a joint may improve function by putting the joint in a better position without bringing the patient back to a normal state. With orthotic fitting, lack of motion at a joint such as the ankle results in an increase in energy consumption and an alteration of gait symmetry, compared to normal. But without orthotic fitting, a poor joint position such as ankle equinus may make the gait even more inefficient and asymmetric. Similarly, ankle-foot orthoses (AFOs) thought to work through changes in spastic reflexes are shown with gait studies to be less effective than hinged AFOs in children with cerebral palsy. Other studies also showed that design of orthoses can be evaluated. In hemiplegic stroke patients, the use of a cane or the combination of a cane and an AFO were found to decrease energy consumption significantly.
Gait analysis is also improving surgical decision making. One study allowed experienced clinicians to review videotape and clinical examination data before recommending a treatment plan. After reviewing complete gait data, the same clinicians changed recommendations in 52% of patients. A second study showed 39% of planned surgical procedures on children were deemed unnecessary after gait analysis. Evaluation of a patient's gait before and after a procedure can demonstrate the efficacy of the procedure and assist in postoperative care. A newer study demonstrated that the Van Nes rotational osteotomy allowed ambulation in children with proximal femoral focal deficiency at a more energy efficient rate than children with a Syme amputation. Gait analysis can allow objective comparisons of various procedures. This is primarily applicable, however, to those procedures in which the end point is an improvement of function. For example, walking efficiency as determined by velocity statistically is the same for transfemoral amputation and for proximal tibial replacement for tumor, but below normal for both treatments for age. Such a comparison of therapy choices can aid in the surgical decision process or the informed consent process. Procedures such as joint replacement have pain as the primary indication for surgery, although improvement of function is also a desirable by-product.
The results of surgical procedures such as total knee arthroplasty and total hip arthroplasty depend on a whole host of variables, including the surgeons' experience and the patients' age, preexisting disease, cooperation, and motivation. Clinical evaluation of the results of these procedures is at best crude from a functional viewpoint. For example, evaluation criteria include walking distance and the ability to climb stairs sequentially. Although the application of sophisticated gait analysis techniques to total joint replacement is relatively new, normative data on total knee and total hip replacement surgery have appeared in the literature for some time. These data are shown in Tables 1–7, 1–8, and 1–9. To date, gait analysis is unable to settle controversies that concern prosthesis design and are related to the efficacy of one type of cemented hip prosthesis versus another, the efficacy of cemented versus uncemented hip prostheses. Another controversy is the issue of the efficacy of prostheses that sacrifice the posterior cruciate ligament of the knee versus those that preserve this ligament and whether mobile bearing knee designs demonstrate a benefit over nonmobile bearing knee designs. Thus far, variability in the data has prevented demonstration of a clear difference in designs. As advances in prosthesis design and gait analysis continue, improvements in the management of patients with gait disorders will also continue.