Pathomechanics and Classification of Cartilage Lesions, Facilitation of Repair loan M. Walker, PhD, PT'

Joan M. Walker

J oint forces have a high po- tential to promote degener- ative changes in articular cartilage, which initially are subclinical and may not ever produce symptoms. Repeti- tive impulsive loading of

activities throughout the lifespan, but especially in adolescence and young adulthood, has potential conse- quences in later life. An individual who runs or jogs 20 miles per week over 30 years delivers some 50 mil- lion shock waves through his/her body across joint surfaces (65). Work- ers in industry may spend decades standing and moving on unyielding concrete floors; they also subject joint surfaces to potentially damaging loads. This paper will review the knowledge of why articular cartilage breaks down, pathomechanics, how cartilage lesions are classified i n vitro and i n vivo, and its capacity for re- pair, with the primary focus on pro- cedures that impact rehabilitation. Other papers in this issue (Cohen et al, Buckwalter, Gillogly et al) should be consulted for detail on the struc- ture and function of articular carti- lage, its response to injury, and surgi- cal repair procedures.

joint forces have a high potential to promote degenerative changes in articular cartilage. Researchers have not yet developed a material that simulates natural articular cartilage, and replacement procedures have finite lives. In all patients, regardless of diagnostic category, the impact of rehabilitative procedures on the integrity and health of articular cartilage should be a consideration. In this paper, I will review why articular cartilage breaks down, how cartilage lesions are classified in vitro and in vivo, as well as cartilage's capacity for repair and repair enhancement. The primary focus will be on processes and procedures that impact physical therapy. Review sources included common computer-based search instruments and literature in all languages. This research showed that most studies have been conducted on animals, which differ in important respects from humans. Such studies, however, provide guidelines for physical therapists. Unloading and overloading are detrimental to articular cartilage. Research indicates value in controlled, progressive regimes that alternate load and non-load conditions.

Key Words: articular cartilage, pathomechanics, rehabilitation

' Professor, School of Physiotherapy, 4th Floor, Forrest Bldg., Dalhousie University, 5869 University Ave., Halifax, NS, Canada B3H 31.5

Pathomechanics or What Causes Cartilage to Break Down?

The extracellular matrix, synthe- sized by chondrocytes, gives articular cartilage its specialized loading and mechanical properties, such as the ability to withstand both static and dynamic loading stresses. The ab- sence of vascularity, however, is a crit- ical factor in articular cartilage's in- ability to respond to injury with the typical stages of inflammation and seriously impairs its repair capacity. The chondrocyte is at particular risk because of this location in an avascu- lar matrix, which creates a long tran- sit route of its nutrition from subsy- novial capillaries through synovial lining tissues, synovial fluid, and carti- lage matrix. Matrix destruction re- sults from an imbalance between anabolism (synthesis) or catabolism (breakdown).

Mechanisms of articular cartilage breakdown involve multiple factors

and imbalance between extracellular matrix degradation and synthesis. Radin (82) considers repetitive im- pulsive loading to be a major factor. Other factors are stress deprivation (immobilization, bed rest, weightless- ness), excessive loading (extrinsic as in carrying heavy loads and intrinsic as in obesity, which is a statistically significant factor in osteoarthritis of the knee joint), malalignment as a result of developmental etiologies (for example, patella alta (49), tibia1 vara, joint surface incongruity) and joint instability (cruciate ligament trauma, menisectomy, generalized ligamentous laxity)].

The type of stress to which carti- lage is regularly subjected appears to condition the structural and func- tional properties of articular cartilage (92). There is also evidence, from canine studies, that the biomechanical response to stress, such as long dis- tance running, is highly sitedependent

Volume 28 Number 4 October 1W8 JOSPT

or sitespecific (9), as the response var- ies between different zones of articular cartilage (superficial, deep) and in prominent weight-bearing areas com- pared with the joint periphery.

The type of load and manner of loading may be more important than the actual load on joint surfaces (82). Repetitive impulsive loading is detri- mental to both articular cartilage and subchondral bone. The more rapid the loading, the higher the tensile and shear rates and the greater the potential for damage. Repetitive im- pulsive loading produces subclinical microfractures in subchondral bone, leading to cumulative damage. Repet- itive impulsive loading may activate

Mechanisms of articular cartilage

breakdown involve multiple factors and imbalance between extracellular matrix

degradation and synthesis.

-- -----.- - - - . .-------

the secondary ossification center, which gradually thins the articular cartilage layer and increases the shear forces in the deeper layers. The tidemark advances and the noncalci- fied layer thins. There is loss of pro- teoglycans, surface fibrillation, and vertical and horizontal splitting of articular cartilage. Splitting and fibril- lation of the cartilage surface pro- duces articular debris in the synovial fluid, which further irritates the syno- vial lining tissues, maintaining a chronic inflammatory state. Catabolic enzymatic activity contributes, and, eventually if the process is continued, cartilage is abraded, the subchondral bone is exposed.

Degenerative cartilage exhibits extensive damage to type I1 collagen fibers, eventual unwinding of the he- lix, and the appearance of collagen fibrils (81). This process appears to commence in the superficial layers and progresses to the deep layers. While cartilage proteoglycan loss can be reversed, loss of the structural in- tegrity of cartilage and its collagen framework results in irreversible dis- integration of the cartilage (22,44). Repair is with fibrocartilaginous tis- sue in which type I collagen, rather than the normal type 11, predomi- nates. This has negative conse- quences as it may facilitate calcifica- tion of repair tissue; only type I1 collagen, now decreased in quantity, has the capacity to block the deposit of hydroxyapatite crystals required for calcification. As will be addressed in a later section of this paper, there is some evidence that given time type I1 collagen may replace the type I collagen formed initially with insult to healthy (that is, not degenerative) cartilage. This process may take up to a year and may be incomplete (6,28,71).

Collagen turnover increases in conditions which increase in preva- lence with aging, such as hypopara- thyroidism and metastatic diseases of bone (81). The degradation of colla- gen under these conditions or dis- eases provides the body with a source of amino acids. It is hypothesized that, in these conditions, articular cartilage is more vulnerable and should not be subjected to heavy loading stresses. It is suspected that mechanical forces, or enzymes, may degrade the "glue" that binds the collagen fibers into a network and be a contributing factor in pathogenesis of cartilage degeneration leading to osteoarthritis ( 15).

Influam of t a p m t u r ~ The enzy- matic processes in cartilage break- down involves both chondrocytes and synoviocytes; both can produce deg- radative enzymes and protease inhibi- tors. Matrix pH and physical factors, such as temperature, influence the

enzymatic activity. Collagenase is more active at high joint tempera- tures (36" vs. 33°C) (68). If cells are even mildly heated, they synthesize heat shock proteins, which are found in synovia from arthritic joints. Heat shock proteins are molecules pro- duced in response to various stimuli with an ability to bind to and influ- ence the intracellular function and distribution of other proteins. They appear to provoke a cellular stress response. It is believed that the tem- perature of an inflamed joint has the adverse effect of inducing synthesis of these proteins (36.1 19). In rheuma- toid arthritis, serum antibodies to these proteins are present. Such data support the use of cold over heat in acutely inflamed joints.

Matrix cornponats The mechani- cal resilience of cartilage matrix is dependent on the chondrocyte's abil- ity to maintain its constituents in nor- mal proportions. The extracellular matrix is composed of collagens, pro- teoglycans, noncollagen proteogly- cans, and water (6540%). Major component5 of proteoglycans are the polymers, chondroitin sulphate and keratan sulphate, members of the group of polysaccharides called gly- cosaminoglycans or mucopolysacchar- ides. The ratio of chondroitin sul- phate and keratan sulphate varies with age, site, and between individu- als (22). The ratio of chondroitin sulphate and keratan sulphate is a factor in the fixed charge density which controls the distribution of charged solutes and, hence, osmotic pressure. Keratan sulphate concentra- tion has been shown to increase with aging and, in osteoarthritic cartilage, decrease with immobili7ation, and may contribute to cartilage pathology (1 4,22,47,64,lO2). Increased propor- tion of keratan sulphate vs. chon- droitin sulphate appears to be a neg- ative change impacting matrix stiffness, its ability to bear loads, and dissipate forces through cartilage.

Inflnmmation The typical inflam- matory response to injury or pathol-

JOSPT Volume 28 Number 4 October 19%

ogy is more visible in the most vascu- lar joint tissue, the synovial lining tissue. Synovitis may result from a variety of stimuli and creates an envi- ronment that is hostile to articular cartilage. Such stimuli can include substances normally foreign to the joint space, such as cartilage wear particles, matrix molecules, immune complexes, hemosiderin (an iron blood corpuscle pigment), implant wear particles (ie., from artificial l ip - ments) , hydroxyapatite (released from bone matrix), and surgical chemical agents (ethylene oxide, glu- taraldehyde) (85). The absorptive synovial intimal cells take up such substances which have been shown to persist indefinitely in the intimal layer. This process, however, may cause the intimal cells to secrete s u b stances that are harmful to articular cartilage and provokes a chronic in- flammatory reaction.

The enzymatically degraded carti- lage releases proteoglycans, which may also activate the synovial cells (so termed "suicide" reaction, where car- tilage hastens its own destruction). A vicious cycle develops with synovial intimal cells releasing more collage- nase and proteinases, cytokines, and interleukin-1 , which further weakens the cartilage and enhances mechani- cal damage (16,80,94). Type B syne vial cells may be primarily involved in this reaction. Collagenase, gelatinase, and stromelysin can digest the collag- enous lattice within the articular ma- trix (26). Type A synovial cells are believed to release cytokines (chemi- cal messengers), such as interleukin-1 and prostaglandin E2, which may play a major role in the perpetuation of synovitis (39). Interleukin-1 is an inflammatory mediator that can cause chondrocytes to decrease ma- trix synthesis and resorb their sur- rounding matrix. Synovial cells are also activated by the proteoglycans released from enzymatically digested cartilage and may influence the im- mune 'response (39).

FIGURE 2. Intra-articular pressures (IAPs) recorded at a range ofjoint angles during active (0) and passive (@) positioning of the knee before and ah'er injection of 10 ml of dextran into the joint 10). Coincident intra- articular pressures shown for both active and passive positioning. Note that in the effused knee intra-artic- ular pressures are lowest at about 30-404 (From Bax- endale et a1 (131, reprinted with permission.1

FIGURE 1. lateral views ofan interphalangeal joint in AJ extension and BI flexion. Redundant svnovium (shown schematicallv) gathers above the superior mar- gin in extension and below the inierior margin in flexion. (From Simkin (941, reprinted with permission).

Intra-atticular pwssurp Clinicians should recognize that effusion compre mises joint stability because the intra- articular pressure that is normally neg- ative becomes less negative or achieves a positive value. A patient with a knee effusion will adopt a posture of about 30-40" of flexion when cavity tissues are under the least tension (Figure l ) , (94). Motion away from this position, particularly active motion, will increase intra-articular pressure and the poten- tial for pain. As strain is increased on

supporting structures, such as l i p - ments, the joint is less stable (Figure 2) (13, 40).

Changes in either the synovial lining tissues or the articular cartilage impact the normal lubricating mech- anisms of the joint. The boundary lubricating ability of synovial fluid becomes more important when syn* vial fluid loses its viscosity due to in- jury or disease. Loss of cartilage corn- - . pliance, that is its ability to deform under loading, increases with aging and reduces the ability to maintain fluid film by the squeeze film or elas- tohydrodynamic mechanisms (both have fluid extruded from unloaded areas of the cartilage; elastohydrody- namic includes cartilage compliance in the model). Theoretically, such conditions lessen the ability of carti- lage to withstand static and/or high loading. Static or high loading are common states in the knee joint com- plex, as in such activities as ascending or descending ramps and stairs and rising from lying. Many sports activi- ties are associated with very high loading conditions.

Effect of Unloading and Immobility-Rest

Articular cartilage is sensitive to abnormal loading. When unloaded

Volume 28 Number 4 October 1998 JOSPT

and deprived of mechanical stimuli, there is rapid deterioration of bio- chemical and mechanical properties; cartilage becomes less stiff and is more vulnerable to injury. Mechani- cal forces, such as loading, may be more important than motion in maintaining cartilage properties; joint movement in the absence of loading was associated with cartilage atrophy (76). and such studies strongly s u p port this contention.

Unloading Studies of denenated and immobilized joints support the suggestion that chondrocytes are sen- sitive to mechanical stimuli (99). In disuse atrophy, such as in paraplegia (27) or poliomyelitis (30) and dener- vation (3), chondrocytes showed in- creased synthetic activity with me- chanical stress and lower activity with decreased compression on the load-

Loading may be more important than motion

in maintaining cartilage properties.

ing surfaces. A decrease in the thick- ness of articular cartilage (atrophy) and chondroitin sulphate observed in these conditions was also reported, with a decrease in glycosaminoglycan synthesis, in studies using animal models and hind paw amputations (46,105). Partial amputation of a limb decreases compression while allowing motion in the remaining joints. Under these conditions, adult rabbits may show changes within 1 week; however, in dogs, similar changes may take 11 weeks to be de- tectable (35,50). Cultured chondro- cytes reportedly respond to mechani- cal stress in a comparable manner to articular cartilage in vivo (109). Colla- gen synthesis, but not the total amount of collagen, may be en- hanced. Effects of unloading on col- lagen may occur in all collagenous

tissues. Fiber-fiber distances and lu- brication between fibers is decreased. This hinders fiber-on-fiber gliding and enhances cross-linking and adhe- sion formation (1 12). The collagen arrangement is more random. Fibers may be thicker, as fewer fibers are laid down without regard to the tis- sue's mechanical requirements (1 15).

Variability in result? of different studies relates in part to problems in the use of animal models, such as differences in species and age and variability in methods, for example, the position of immobilization, the degree of compression of the joint surfaces, and the extent that motion is limited. Reduced loading studies support the view that maintenance of healthy joint surfaces requires a regu- lar program of loading. Encouraging individuals of all ages to walk regu- larly should be beneficial.

Inflamed joints are subjected to unloading and temporary immobiliza- tion for short time periods. As rest is a basic principle of inflammation management, whatever the cause, the negative effects of immobilization of joint tissues produce a dilemma: how to determine the balance between the period of rest to decrease inflam- mation, yet not incur further func- tional loss? Effusion reduction (thermal applications) with anti-in- flammatory drug therapy are impor- tant components of treatment. If car- tilage that has been atypically unloaded is then loaded, there may be "gross functional failure of matrix" (75). This appears to be related to a defect in the ability of proteoglycans to reform the hyaluronic acid-binding region of the aggregates. Impairment of the proteoglycan matrix may make articular cartilage more vulnerable to injuries if exposed to sudden, heavy loading. Studies on animals suggest that such changes may be reversed by a gently graduated program; how- ever, restoration of healthy structure and characteristics is uncertain (101). Therapists should recognize that fol- lowing immobilization or unloading (prolonged rest), articular cartilage is

less stiff and less capable of tolerating high loads, loads that normally are within the physiological capacity of healthy cartilage. These results sug- gest that there is a need for pro- grams of graduated weight bearing and activity after particularly ex- tended periods of non-weight bearing and/or casting.

Immobilization Immobilization should prevent motion. Depending on the joint position, either an in- crease or decrease in loading to areas of the joint surfaces may occur. This limitation in motion results in degen- erative changes that are similar to that seen in osteoarthritis and may result in cell necrosis (4,35,37,42,46, 104). When clamps are used to pro- duce immobilization (knee joint in small animals), clamping may also have caused compression of the artic- ular cartilage (29,33,43,90,107,108). Cell death occurs with rigid, sus- tained pressure at focal points of the articular cartilage. Presumably, there is interference with cellular nutrition which is dependent on the cyclic compression and expansion of articu- lar cartilage that occurs with normal intermittent loading (99). Simple dif- fusion of small solutes may be unaf- fected by cyclic loading (106). Perme- ability of cartilage decreases with compression of loaded areas (41) (Figure 3). Maroudas indirectly dem- onstrated that permeability of carti- lage correlated inversely with hydra- tion (66). Large deformation of cartilage will regulate the manner and rate of fluid transport; the more rapid the applied load, the more rapid the efflux at the edges of the loading surface, as well as movement of free fluid from superficial to deeper layers. It must be recognized that, in healthy cartilage compared with pathological cartilage, under any condition, fluid movement is minimal because most fluid is bound or con- strained by the collagen lattice frame- work of the extracellular matrix.

Simple immobilization without compression also causes atrophy and

JOSPT Volume 28 Number 4 October 19%

C L I N I C A L C O M M E N T A R Y - - . . . - -- - . - -- . - -

1 LOAD

- subcho~dral bone

FIGURE 3. Simplified diagram of fluid motion in artic- ular cartilage under loading. unc = Uncalcified layer. Fluid motion will vary with the size of the load and rate of loading. In loaded areas, free fluid will move into deeper layers until limited by the impervious tidemark zone; free fluid also moves sideways into less or unloaded areas and may efflux at the surface in minute amounts. In healthy cartilage, most matrix fluid is bound and not free to move.

thinning of the articular cartilage but no change in cell density (41,98). Other changes include a decrease in glycosaminoglycans and chondroitin sulphate compared with the nonoper- ated or immobilized side (25), whereas intermittent compression of articular cartilage causes the reverse (75,109,111). With the atrophy of articular cartilage that has been ob- served in immobilization, there is also an enhancement in collagen syn- thesis (greater collagen fiber diame- ters) but not in the total amount of collagen (51,78,113). Tammi et al (100) used splints to immobilize dogs and showed a 53% increase in colla- gen production. In comparison, there was only an 11-13% increase in the dogs who ran on a treadmill for 15 weeks. Variability in results of dif- ferent studies may relate to age of the animals, type of immobilization, and that some areas of the cartilage joint surface are subjected to in- creased loading while other areas are unloaded, dependent on the position of immobilization. The inconpi ty between the femoral and tibia1 con-

dyles, as well as the varying contact between the patella facets and the femoral surface dependent on the joint position, makes this more likely in the knee joint.

When immobilization is an essen- tial component of patient manage- ment, as for fracture management or following surgery, clinicians should devise exercises that will provide loading to immobilized joint surfaces. As yet, there are no studies to dem- onstrate whether this approach would result in "stronger" cartilage.

Increased Loading--Compression

Cartilage degeneration is associ- ated with excessive mechanical load- ing (such as repetitive impulse load- ing) that produces fatigue fractures, microscopic cracks in and around osteons that involve individual trabec- ulae, and, over time, reduces the strength and stiffness of bone. If the damage is excessive, resorption will exceed deposition and failure will occur at the macroscopic level, as in crush fractures and contour changes in joint surfaces. These changes in bone place further stress on articular cartilage. Research has not yet estab- lished if such microcracks can trigger a cellular response to bring about repair of the damaged matrix.

There have been a number of experiments in which the joint sur- faces have been abnormally loaded. Non-weight bearing on one limb cre- ates an increased and altered load on the weight-bearing side. Studies in which the opposite limb joint(s) was used as a control(s) then have imper- fect control tissue. Backpacks may be put on the animal, or weight gain may be simulated as in experiments performed by Simon et al with rats (95). Severe excessive loading experi- ments have employed the "drop tower" procedure to subject articular cartilage to high stress levels. These studies have shown that chondrocytes can survive up to a 30% compressive strain, but if the stress exceeds 250 g/mm2 and produces a strain that is

greater than 40%, tissue death results (84). This type of impact loading re- sults in a loss of proteoglycans, fibril- lation of the surfaces, and cell death (83). When loaded tissue extrudes interstitial fluid, it becomes less hy- drated, resulting in an increase in the proteoglycan concentration which then, in a feedback mechanism on chondrocytes, inhibits its own synthe- sis (99). In vitm studies have revealed that static compressive pressures con- sistently inhibited proteoglycan syn- thesis in cartilage (77). Review of the effects of exercise suggests the re- sponse to cyclic compressive loading is more variable.

Excessive loading may occur in various sports, especially those that subject joints to impact and/or tor- sional loading, such as in ice skating with repetitive jumps and landings on a hard surface. Studies (7,56,59) sug- gest activities such as jumps should be interspersed with much longer periods of nonimpulsive loading to allow reconstruction of the cartilage following severe compression. There are implications for both rehabilita- tion and training routines. Loading activities, such as treadmill walking and step machine usage, should be followed by a period of exercise which unloads the previously loading joints: swimming or trunk or arm exer- cise in a different position, ie., lying.

Effect of Exercise

While many therapists believe that joint motion improves circula- tion and, thus, nutrition to articular cartilage, van Kampen and van de Stadt (109) following review of in v i v o and in v i t r o studies concluded that nutrition plays only a minor role in the processes involved in articular cartilage's adaptation and response to under or overloading. The role of exercise in cartilage breakdown is unclear. Although some investigators have shown negative changes follow- ing exercise (10,11,24,57,110,113,114, 1 16), positive effects of exercise have been demonstrated (48,55,60,79,86-

Volume 28 Number 4 October 1998 JOSPT

-. - - - - - - - C L I N I C A L C O M M E N T A R Y

, - -

88). The type of exercise influences the response. There were more de- generative surface changes in the femoral heads of rabbits with "sud- den maximal" exercise compared with submaximal running (24). Sev- eral in v i t r o studies have shown that the chondrocytic response to loading, particularly cyclic stresses, may vary with the length of the treatment. Palmoski and Brandt (75) subjected cartilage plugs to static and cyclic stresses that were equivalent to about 1.5 times body weight. They demon- strated a positive response, a 38% increase in synthesis of glycosamino- glycans with a cycle of 4 sec on/ 1 l sec off, but also a negative response, a decrease with a cycle of 60 sec on/60 sec off (75). This varied re- sponse has been shown both in cell cultures and in chondrocyte cultures (79). The response varied with the frequency of the pressure, duration of the application, and the state of the cells. Application for 1.5 hours produced a strong inhibitory reac- tion, whereas a 20-hour exposure produced a positive reaction, stimula- tion of sulphate incorporation (77). Response variability may relate to the observation that it may take 1.5-2 hours for complete aggrecan mole- cule synthesis, a finding which physi- cal therapists should consider in planning therapy because aggrecans contribute to cartilage's ability to withstand loading. It suggests the use of alternating periods of unloading and loading and, in the early stages of rehabilitation, unloaded periods should be longer than the loaded periods.

When cartilage explants were ex- posed to cyclic stress of two different frequencies (high: 2 sec on/2 sec off; low: 60 sec/60 sec o m , there was a decrease in protein and proteogly- cans synthesis with low frequency and static loading, whereaq, in the high frequency cycle, there was a stimula- tory effect on protein and proteogly- can synthesis (61). Unloading re- stored the synthesis to the preloaded levels. Results of these studies suggest

that in the early stages of rehabilita- tion unloaded, quicker movements should be used. When static loading is introduced, time in unloaded con- ditions should follow and be of a longer time interval. Research is vi- tally needed to determine how differ- ent exercise regimes affect in v i v o cartilage.

It should be noted that in v i t r o studies expose the tissues to only one type of stress, with forces that are mostly lower than the physiological levels experienced in v ivo , such as at heel contact and toe-off in the gait cycle. Studies on knee joints of quad- rupedal animals record forces in an individual joint that would be higher in bipedal humans. Quantitatively, contraction of muscles may result in greater forces across articular carti- lage than does normal weight bear- ing (16). For instance, while 4-5 X

body weight (BW) may be transmit- ted through the knee in walking, the force may be as much as 10 X BW in squatting (97). Similar reaction forces are estimated to occur in gym- nastics (rake-off vertical forces = 3.4- 5.6 X BW) (18). High transarticular forces can be predicted for all j u m p ing. Studies do show that joint sur- faces can tolerate higher dynamic than static forces. It is also important to recall that loading is not uniform in the human knee joint and varies between condyles and compartments (19,23,92).

As microtrauma is cumulative, it is important to remember that equip ment can play a major role in d a m p ing joint forces. Is the footwear worn appropriate to the activity and in sound condition? A new mat in gym- nastics was shown to reduce tibia ac- celerations from 50 to 10 g and hip accelerations from 20 to 8 g (18). Sheep that had walked for long dis- tances 4 hours/day for 12-30 months on concrete floors demonstrated negative effects in knee articular car- tilage, but these changes were not seen in those that walked on wood chips (83). If practice and perfor- mance jumping times are summed

STRESS Excessive

FIGURE 4. Tammi et al's hypothesis of the effect of intermittent physical stress on the biological properties of articular cartilage. They suggest that atrophy or injury occurs in articular cartilage due to suboptimal physical stresses, while stresses of normal use within physiological limits stimulate the tissue to develop improved biological properties. Increasing the stress beyond this physiological boundary causes a progres- sive deterioration of its biological properties. This same rule may also be applicable to other connective tissues of the locomotor system. (Adapted from Tammi M, Paukkonen K, Kiviranta I, luwelin 1, SGmanen A-M, Helminen HI: Our hypothesis of the effect of intermittent physical stress on the biological properties of articular cartilage. In: Helminen HI, Kiviranta I, Saamanen A-M, Tammi M, Paukkonen K, lowelin / (eds), joint Loading, Chapter 3, p 82. Bristol: Wright, 1987, with permission.)

from sports, such as parachuting, hurdling, or ice skating, it is obvious that these activities result in high re- petitive impulsive loading to joint surfaces.

Tammi et al (101) reviewed stud- ies of various animal species involving "enhanced loading," which have pro- duced mixed results as to whether running exercise injures or does not injure articular cartilage. One factor in the variability of results may be the rate at which the program is com- menced ( 101 ) . A conditioning period seems important. The positive effects on articular cartilage of enhanced loading in the form of moderate run- ning (eg., 4 km/day) are reversed by strenuous running (eg., 20 km/day). Whether the response to nonstrenu- ous exercising in human athletes is similar to that of animals, mainly dogs, is not yet known. The limits of the "physiological range" (Figure 4) and the response of mature articular cartilage in v i v o needs to be deter- mined. Study results suggest that ex-

JOSPT - Volume 28 - Number 4. October 1WU

C L I N I C A L C O M M E N T A R Y -. --. - ---..--- --.-- - - --.--- -..--.---.--up-

ercise involving loading should be progressed gradually.

In summary, cartilage breakdown has multifactorial causes and the mechanism is not fully understood. Abnormal loading, be this excessive or reduced, whether from sustained static positioning as in immobiliza- tion or from activity/exercise or sport, changes cartilage properties and makes it more vulnerable to in- jury. Comparison of results from dif- ferent studies requires a valid and reliable system of classifying cartilage damage which is used by different investigators.

Classification of Defects

Currently, we lack a reliable, fea- sible method of monitoring the state of joint cartilage in vivo. It is not fea- sible to biopsy articular cartilage. Even in osteoarthritis, "the criteria for diagnosis and classification on the basis of clinical and/or radiological presentation of even the overt stages of osteoarthritis are under dispute" (63). Lohmander noted that some sets of criteria in use neglect to even mention damage to the central issue, that of the joint cartilage itself (63). For example, in osteoarthritis of the knee, radiological criteria include the presence of osteophytes (spurs), nar- rowing of the joint space of one or more compartments, especially with weight bearing, subchondral sclerosis, and cysts (5). Such a list, however, does not include cartilage damage. Similar to goniometry, intraobserver error on reading of radiographs var- ies at different joints, and two read- ings by one observer has closer agree- ment than two readings by separate observers (5.53). Advances in radiol- ogy, such as magnetic resonance im- aging (MRI), may enable in vivo de- tection of cartilage detail, especially its response to loading.

Classification of tissues has a key role to play in the interpretation of results among groups of samples, be- tween patients and between animals and humans. There is a need to de-

Author Material Grades ReliabiTi Bennett et al (14) Human postmortem 1 t uneven, granular surface, furrows NR

2 + also fraying, splitting, some erosions 3 t extensive ulceration 4 t exposure of subchondral bone

Kiss et al (54) Human necropsies 0 intact system NR P t moderate alteration, color, softness P+ t moderate (same items) A t partially destroyed, osteophytes, bone exposed

(<2 cm2) A t t bone exposed over whole surface

Mitrovic et al (70) Autopsy human 0 no lesion NR 1 slight fissure 2 severe fissure 3 slight deep ulceration 4 large ulceration (>25% cartilage surface with

subchondral bone exposed)

NR = ,Not reported.

TABLE 1. Grading systems for postmortem material at the macroscopic level.

velop a diagnostic/disease progres- sion method that will enable integra- tion of clinical findings with b i e chemical, biomechanical, etiological, and histological evidence. The follow- ing discussion of grading systems to classify cartilage defects focuses on histological and postmortem mate- rial; complex rating/classification systems used clinically for knee disor- ders will be covered elsewhere in this special issue. Classification or grading systems that have been used on au- topsy or histological materials are presented in Tables 1 and 2 (52,60, 64,69,70,93,116). Most are descriptive at best; semiquantitative and reliabil- ity rarely has been given. Most lack adequate definition between grades, while sometimes grades appear iden- tical. Some of the differences be- tween reported systems relate to dif- ferences between studies, in which only gross examination of the surface was conducted, to studies encompass- ing gross examination, histochemis- try, and morphometrics. Gross exami- nation may include Indian ink applied to the cartilage surface which allows identification of smooth to deeply fibrillated cartilaginous sur- faces. Grading systems can be used with this method, which is regarded a$ an excellent preliminary tech- nique, even with only slightly abnor- mal surface cartilage (73).

Mankin et a1 in 1971 (64) (Table 2) provided the most indepth hist* logical classification based on com- bining scores for several different parameters, which can then be grouped for an estimate of disease/ damage severity (67). Mankin et a1 used four criteria with subcategories for three of the criteria. The compos- ite score ranges from 0, completely normal, to 12-15 (almost completely disrupted). This system has been used most frequently. The limitation of the Mankin et a1 system is that it is not known if graduations, as mea- sured by the composite score, are related in a linear fashion to the dis- ease process (73). For example, a smooth surface or partial clefts may be strongly related to say, the Safra- nin staining score.

The absence of reliability testing is a weakness of many reports, and it is obvious from use of systems re- ported in the literature that most probably would have unacceptable reliability. Jurvelin et a1 (52). who reported reliability of their grading system (Table 2), noted that "qualita- tive methods are the basis for quanti- fication, but are not sufficient in themselves." They advocated classifi- cation of structural details present and the need for determination of the proportion of each type/category across the entire surface. Their sys-

Volume 28 Number 4 October 1998. JOSPT

C L I N I C A L C O M M E N T A R Y ." ..-------.,-..-. . . . . . . .. . - -. . . - .

' ~;lt);dr- Material Grades Reliability

Lanier (60) Mice 1. Minimal roughening, fibrillation, and splitting, extending to but not None reported involving calcified cartilage

2. Degeneration extending into calcified cartilage, loss of surface not extending into subchondral bone

3. Lesions involving subchondral bone, thickening, and ebumation of cortex

Mankin et a1 (64) Human 1. Structure, six subcategories: Two observers, samples .1 Normal 0 repeated for validity, .2 Surface irregularities 1 variation estimated at .3 Pannus and surface irregularities 2 < 5 O/O

.4 Clefts to transitional zone 3

.5 Clefts to radial zone 4

.6 Clefts to calcified zone 5

.7 Complete disorganization 6 2. Cells (three subcategories: normal = 0, diffuse hypocellularity = 1,

cloning = 2, hypercellularity = 3) 3. Safranin-0 staining (four subcategories: normal = 0, slight reduction = 1,

moderate reduction = 2, severe reduction = 3, no dye noted = 4) 4. Tidemark integrity (one subcategory: intact = 0, crossed by blood vessels = 2) Note: All grades with a 0 = normaliintact.

Jurvelin et al (52) Rabbit patella Scanning microscopy N = 10 1. Evenness of surface [six subcategories (even, slightly uneven, hillocky, Mean difference between

folded, fibrous, unc1assified)l two investigators of 346, 2. Roughness of surface [six subcategories (wen, slightly rough, rough, knobby, variation 5%; one person

leafyktriated, unclassified)] did all evaluations 3. Splits of the surface [six subcategories (intact, superficial splits, deep splits,

unclassified)] plus a com uterdrawn map of joint surfaces Total and mean area of each c k s Minimum, maximum, and relative area (%) of each class compared with whole

cartilage surface

Walker (1 16) Rat knees 1. Surface only fraying, fibrillation, small defects in uncalcified cartilage layer r = 0.96 (UCL) with abnormal staining reaction

2. Surface defects extended into UCL, cracks, holes in UCL, observed in 10-20% of sections

3. Prominent defects, multiple defects within a single section, defects extending from surface to tidemark, present in >20% of sections, marked abnormality of cartilage

Shortkroff et a1 Dogs, chondral and Grades for Masson-trichrome stained sections (93) osteochondral 0 = No positive staining

defects + = Faint staining (<than normal) + t = Moderate staining (same as normal) t t t = Intense staining (more than normal) Safranin-0 stained sections: 0 = No positive staining t = <25% staining ++ = 25-50% staining +++ = >SO% staininr!

None reported

TABLE 2. Grading systems, light and scanning electron microscopic methods.

tem was supplemented with a com- puterdrawn map of the entire sur- face. Such an analysis is claimed to enable analysis of alterations on a joint surface with greater accuracy than had been previously achieved. Reports of variation in structural and biomechanical properties by site within a joint (9,51) indicate the im-

portance of sampling across the en- tire joint articular surface and of both cartilage and subchondral bone, if analysis and classification is to re- veal valid data.

Regardless of the classification system used, it is vital that investiga- tors consider the potential artifact- inducing effect of specimen prepara-

tion. Variability in samples, in the chemical composition and morphol- ogy, may change in response to p rep aration techniques. Motion analysis ideally should precede tissue analysis in order to properly determine how the joint is used and the distribution of loading during normal joint mo- tion. Macroscopic or microscopic

JOSPT Volume 28 Number 4. October 1998 223

analysis should be accompanied by biomechanical analysis, as the ability of the tissue especially in the knee to withstand loading may be more criti- cal than its morphological appear- ance.

Physical therapists should apply a critical evaluative approach to the literature and ask the following ques- tions:

Was the sample defined? Were all animals/subjects/ samples accounted for? Were controls reported? Were these different subjects or oppo- site joints? How were animals assigned to study groups? Were investigators, assessors masked (blinded) as to study group? Were tissues evaluated by one or more assessors? Was reliability (consistency) re- ported? Was the position ofjoint surfaces during investigation, tissue fix- at ion, and sampling clearly shown? Was location of cartilage samples precisely described, were sites se- lected at random? Were tissue preparation artifacts controlled, discussed? Use of a simple system, where

Yes = 1, No = 0, should assist read- ers in critical appraisal of this litera- ture.

Capacity for Repair

The ability of any tissue to ex- hibit the typical protective response to injury or disease, of inflammation, is strongly related to its vascularity. In joint dysfunction, synovial lining tis- sues invariably demonstrate a degree of inflammation while hyaline articu- lar cartilage being avascular does not. Recall that a histological criterion for abnormal mature articular cartilage is the presence of blood vessels crossing the tidemark (64). With inflamma- tion, there is cellular infiltration which, in acutely inflamed tissues, is dominated by polymorphonuclear

leucocytes, wherea? in chronic in- flammation, mononuclear cells, mac- rophages, and lymphocytes dominate.

Neovascularization (angiogenesis) appears necessary to sustain chronic inflammation (38). Angiogenesis is the primary vascular response in chronic inflammation and provides fresh monocytes and other inflamma- tory cells. Fibrosis occurs in chronic inflammatory states and suppuration where a bacterial infection is present. The fibrosis phase may convert the proliferative fibro-fatty connective tissue formed in the inflammatory process into adhesions, which mature into strong scars, particularly if the part is immobilized. Such scars may result in significant joint contractures with consequent loss of mobility.

Healing with restoration of the original tissue is dependent on the intrinsic ability of the tissue to regen- erate and the complexity of the tis- sue. As true regeneration does not occur in articular cartilage, there is replacement by fibrous tissue which organizes into a firm scar and fre- quently lacks the important charac- teristics of articular cartilage. Hyaline articular cartilage has a very limited and imperfect repair mechanism, particularly where the defect only involves articular cartilage and does not extend into the subchondral bone plate (58,96). The isolated liv- ing chondrocyte is not, as formerly thought, an effete cell. Chondrocytes constantly change shape, put out, and withdraw pseudopodic processes. If in vitro cells from the superficial layer are transplanted to a deeper layer, they will change their shape and become more rounded (99).

Bone, a tissue with high regener- ative capacity, is very responsive to stress, as from a change in loading conditions. With progressive damage to articular cartilage and exposure of collagen fibers, there is altered load- ing on the subchondral bone which responds by remodeling. This may result in thickening (sclerosis) of tra- beculae and in the formation of spurs (osteophytes) as seen in osteo-

arthritis. These bony changes are en- hanced by degradation of articular cartilage or detachment of cartilage secondary to its splitting. Osteophytes may provide a buttress effect, an at- tempt by the body to limit further excessive loading of damaged carti- lage.

The complex processes involved in cartilage metabolism are depen- dent on many factors, including nor- mal cartilage nutrition in which mo- tion and loading play a vital role. There is no apparent difference in repair capacity of different levels of articular cartilage, that is, superficial or deep levels; only defects involving the subchondral plate may show re- pair with articular cartilage-like tissue.

The subsynovial microvasculature and lymphatics play a large role in cartilage nutrition, providing nutri- ents and the normal clearance mech- anism vital for repair to occur. Effu- sion signifies imbalance between microvascular permeability, Especially to proteins, and the normal clear- ance mechanism of the lymphatic outflow and is shown by increased protein concentration in synovial fluid. Effusion is common when in- flammatory processes involve joint tissues. Damaged tissues, particularly synovial lining tissues, are infiltrated with leucocytes; there is an influx of neutrophils via increased blood flow and time-related chemo-attractants. Neutrophils can cause further dam- age secondary to release of proteo- lytic enzymes and 0'-free radicals. Platelets release protease inhibitors which serve to limit tissue damage. It should be recognized that both chon- drocytes and synoviocytes can pro- duce degradative enzymes. A vicious cycle is initiated where the reaction may cause further damage.

Regular motion may adequately clear the leaking protein but rest, sleep, or immobilization allows the synovial lining tissues to become edematous; the individual experi- ences pain and stiffness. Even a once daily movement through the available range appears to be important in ef-

Volume 28 Number 4 October 1998 JOSPT

fused joints, although welldesigned studies remain to be conducted.

A5 a result of matrix degradation, basic fibroblast growth factor and transforming growth factor-p (TGF-P) normally bound to matrix proteogly- cans are released. Such extracellular messenger proteins are involved both in tissue degradation and synthesis repair. The degradative effects of in- terleukin-1 can be controlled by the insulin-like growth factor and TGF-P. which can stimulate collagen tran- scription and induce synthesis by connective tissue cells of tissue inhib iting metalloproteases and plasmin* gen activator 1. Insulin-like growth factors (I and 11) can stimulate chon- drocyte mitotic activity and proteogly- can and collagen synthesis. Fibrobla. tic growth factors also can stimulate chondrocyte mitotic activity.

Repair Enhancement

In the last three decades, there have been a number of histological, biochemical, biomechanical studies. as well as analytical models, that have documented the response of articular cartilage and/or chondrocytes, both in vitro and in vivo. to mechanical stimuli. This section on repair en- hancement of articular cartilage will focus on studies with implications to physical therapy practice; other a p proaches such as grafting and im- plantation will be covered in other papers in this special issue. Regret- fully, there are few studies reported on the direct effect of physical ther- apy intervention; however, studies in which exercise was employed with animal models have implications for physical therapy and training.

Exer&e/Training Exercise/mo- tion in animals and humans has been shown to cause a swelling of articular cartilage. Prolonged exercise in ani- mals was observed to produce posi- tive effects, such as hypertrophy of chondrocytes, an increase in the peri- cellular matrix, and an increase in the number of cells per chondron, particularly in the radial and transi-

tional zones (34). In rabbits, the su- perficial cells became more spherical and larger for a transient period after brief exercise (31 ). Long-term exer- cise or loading produced more last- ing enlargement. In young rabbits given a nonstrenuous treadmill exer- cise program for 8 weeks, there was an increase in the volume of nuclei but not in the overall size of the cells in the superficial layers. There also was an increase in the quantity of endoplastic reticulum and other cyto- plasmic organelles (79). Guinea pig , after a running program, showed an increase in the size of chondrocytes (86). Helminen et al (47) observed an increase in the number but a de- crease in both the volume and vol- ume density of chondrocytes after "strenuous" running in beagles. Fol- lowing moderate, not strenuous, run- ning by young beagles, this Finnish group of workers reported a positive change in articular cartilage proteo- glycans (87) and no degeneration of the surface after a 15week period (55). Long distance running (up to 40 km/day. 5 day/week for 1 year) by dogs showed depletion of glyco- saminoglycans from the superficial zone of articular cartilage and soften- ing of articular cartilage at the same sites (9-1 1 ) . A1 though these changes did not lead to articular cartilage breakdown, they do resemble those of early osteoarthritis. This group also has observed an increase in artic- ular cartilage thickness, increased cell size, and improved anabolic processes following exercise training in animal models (47,55,79,102).

At this time, concisely summariz- ing exercise effects on articular carti- lage is not a simple task due to the variable and often conflicting study outcomes. This variability relates to the age of animals studied, the time postexercise cessation that animals are sacrificed, and the analysis em- ployed. More recently used quantita- tive histochemical and biochemical analyses provide more detail of ef- fects, especially at the cellular level. The intensity of training affects the

end result. Moderately strenuous ex- ercise can be tolerated and may pr* duce positive cellular changes with- out evident degenerative changes. Study outcomes strongly suggest that loading is more important than mo- tion in the maintenance of cartilage properties.

It now should be obvious that we do not yet know which type of train- ing best improves the ability of articu- lar cartilage to withstand stress and that positively contributes to repair of articular cartilage. Exercise is shown to increase blood flow and perfusion of soft tissues in the canine wrist and knee; however, effects on cartilage are unclear (69). Such data support the use of only minimal activity to preserve range of motion and muscle balance in the acutely inflamed joint and also support the value of vigor- ous exercise in chronic inflamed joints.

While cells of healthy articular cartilage appear to have suppression or loss of the ability to replicate their DNA, when degeneration occurs, as in the early stages of osteoarthritis, chondrocytes appear to recover their ability to divide (45,103). Overall, repair by surviving cells seems absent when a cleft involves only articular cartilage, not subchondkl bone. Salter et a1 (89). over the past two decades, have demonstrated healing of full thickness drill hole defects and free intra-articular periosteal grafts with continuous passive motion with hyaline-like tissue. Results have been less impressive in mature animals. It is not established how newly synthe- sized articular cartilage responds to normal loading over time. The evi- dence, however, is that physical sig- nals, such as motion and load, are important in facilitating cartilage re- pair processes, however imperfect these may be. Data from animal stud- ies also suggest that intermittent rather than continuous mechanical stress induces beneficial metabolic changes in chondrocytes, and a b sence of motion is detrimental.

Approaches to facilitate repair Hyaline articular cartilage is a unique and marvelous tissue which humans have yet been able to replicate. Meth- ods explored to improve repair are biomechanical, biophysical, involve subchondral bone, grafting, surgical implantation, and a combination of these various methods. In this paper, only those that may be a component of physical therapy management will be addressed. Investigators continue to search for methods to stimulate repair of damaged cartilage and pro- vide a near normal articulating sur- face, with some of these methods be- ing employed surgically. If techniques can be employed to simulate the cov- erage of the load-bearing surfaces with a tissue that has loading charac- teristics which permit functioning similar to natural articular cartilage, replacement procedures can be de- ferred. These procedures have a lim- ited tolerance for heavy loading and vigorous wear over the long term. Delay until after maturity is especially important in young patients.

Biowwchanical Variable success from the following biomechanical approaches has been achieved: 1) decreasing the load through the weight-bearing surfaces by use of dif- ferent ambulatory aids (canes, crutch- es); 2) use of continuous or intermit- tent passive motion.

Of particular interest to physical therapists should be the biomechani- cal approach of continuous passive motion reported mainly by Salter and associates (89). Continuous passive motion may be either continuous or cyclic. In adolescent rabbits, Salter and Field (90) demonstrated the healing of full thickness drill hole defects by hyaline cartilage within 4 weeks, with a significant difference in the animals that had been treated by continuous passive motion. The per- centage of articular cartilage in the three experimental groups of rabbits was: immobilized 8%, free motion 9%, and continuous passive motion 52%. O'Driscoll and Salter had simi-

lar results using free intra-articular periosteal grafts (72). These results were less impressive in mature rab- bits. The response of the newly syn- thesized articular cartilage to normal loading over time was not estab lished. Both of these studies showed a combination of approaches, contin- uous passive motion and involvement of the subchondral bone with the use of full thickness drill holes or perios- teal grafts. Recently, simulated repair of articular cartilage matrix was re- ported following chemically induced osteoarthritis when intermittent ac- tive motion followed continuous pas- sive motion (118). Future studies may show that it is the variation in the stimulus of hydrostatic pressure changes that may be more important than the motion imposed.

Coutts et al (28) demonstrated in the mediofemoral condyle of adult rabbits, with a full thickness defect and rib perichondrial grafts, that at 6 weeks nearly normal artic- ular cartilage was present in 55% of animals, and, at 1 year, 82% dem- onstrated type I1 collagen. The shear moduli did not differ between the two groups (those exposed to cage activity alone and those having passive motion followed by cage ac- tivity). Articular cartilage with bio- mechanical properties developed and did not degrade within 1 year after the full thickness defects had been created. The percentage of type I1 collagen, the "right" type, may increase over time (6). Moran et al (71), in a rabbit model with full thickness defects and autoge- nous periosteal grafts, gave either continuous or intermittent passive motion. The tissue was examined with a variety of microscopy a p proaches and histochemistry but with no biomechanical tests. They demonstrated healing by hyaline cartilage that contained type I1 col- lagen, albeit with an irregular orga- nization. This result was better in the animals that had been grafted and better in animals that had con-

tinuous passive motion rather than intermittent passive motion.

Regardless of the methods em- ployed to simulate, induce, or im- prove repair of articular cartilage, the functional behavior of the tissue should be examined biomechanically. Microscopy studies have shown that the repair tissue may have the a p pearance of hyaline cartilage, but, over time, it fails to perform like ar- ticular cartilage. This suggests that an early return to sports which involves repetitive impulsive loading activities (eg., hurdling, running, ice skating with jumps, or an occupation requir- ing repetitive lifting of heavy loads or incline/step climbing) is probably ill advised. A very gradual return to high loading conditions seems pru- dent because it has been shown in animals that the desired type I1 colla- gen takes a year or more to develop. Many professional and amateur ath- letes would probably balk at such a lengthy noncompetitive or low train- ing schedule.

Biophysical Variable effects of electromagnetic fields on cartilage and chondrocytes have been demon- strated mainly with chondrocytes grown in culture plates. No effect of pulsed electromagnetic on extracellu- lar matrix synthesis of chondrocytes in high density cultures was observed (32). However, increased proteogly- can and DNA synthesis occurred with chondrocytes grown in cultures with constant direct current (74). Other studies also revealed a stimulating effect of electromagnetic fields on chondrocyte activity (1,2,17). Two studies have demonstrated enhanced healing of osteochondral defects in rabbits, with experimental animals showing repair tissue like hyaline car- tilage (12,62). The long-term results and biomechanical characteristics of the repair tissue are not known; how- ever, it is reasonable that electromag- netic fields should have a biological effect. Current intensity may be a critical factor. Armstrong et al (8) on growth plate chondrocytes grown in

Volume 28 Number 4 October 1998 JOSPT

electrical fields showed inhibited growth with strong electrical fields but enhancement of proliferation with weaker electrical fields.

Laser Increased synthesis of proteoglycans and collagen was shown following exposure to CO* and Nd:YAGlaser irradiation in or- gan bovine cultures and chondrocyte cultures (1 17). Better repair of super- ficial defects in guinea pig knee oc- curred with lowdose laser energy (25 J) compared with highdose laser en- ergy (125 J) and controls (91). Higher dosages (>75 J) produced cartilage and subchondral bone dam- age. Long-term results were not re- ported, and while the mechanism is unknown, the authors theorized stim- ulation of cell replication mecha- nism. For both electromagnetic and laser therapy, more research is needed.

Many investigators have utilized a combination of different approaches. subchondral bone involvement, graft- ing, and/or surgical implantation. Buckwalter and Mow (21) noted the following limitations of experimental studies. If osteoarthritic joints are used, it should be remembered that the changes in osteoarthritis are not restricted to those of articular carti- lage. Regardless of the technique em- ployed to enhance cartilage repair, it should be noted that animal models used have often been rabbits, guinea pig , or rats, and the small joints that are used do not permit extrapolation directly to the larger joints of hu- mans. Additionally, articular cartilage varies in thickness, cell density, com- position, material properties and function between joints, and species. Further, when animal models are used, there is difficulty in assessing the long-term functional outcome. Few investigators have studied the animals for an adequate period fol- lowing the experimental procedure. Pain relief, often the primary objec- tive of intervention in humans, is dif- ficult to assess in animals. Clinicians should be aware that "the available

evidence indicates that, at a mini- mum, repair tissue must have initial protection from loading, followed by a regime of loading and motion that promotes remodelling (20)" but thus far experimental studies have not defined this optimal sequence of changes in the mechanical environ- ment (21).

Clearly, it is not yet possible to give definitive directions to clinicians with regard to progression of a reha- bilitative program; however, perhaps for a year, repetitive high loads, as in long distance running, should be avoided and, when used, briefly applied.

Implications for Physical Therapy Practice

The knowledge of the response of articular cartilage to loading and to mechanical stimuli is growing. The results of the studies over the last couple of decades appear to chal- lenge the ancient dogma of "rest," since in unloaded joints, rest has proven to be detrimental to the health of articular cartilage. I hypoth- esize that one of the objectives of physical therapy, particularly in ma- ture clients who have been confined to bed for lengthy periods, for exam- ple, in conservative treatment of a fractured femur, should be to pro- vide loading experiences for the un- affected leg to maintain the health of articular cartilage. This could be as simple as a regime of leg pushes against rubber bands or springs and bent-leg and resisted pelvic raises. Probably all health professionals should aim to minimize the unloaded period and, wherever possible, to al- low at least partial weight bearing to provide some mechanical stimuli to the articular cartilage. Additionally, physical therapists must recognize that there is an unknown critical u p per limit to the load cartilage can bear without adverse effects (Figure 4). Following injury and periods of

immobility, cartilage is less stiff and less capable of tolerating high loads.

Lubrication mechanisms should be considered in planning rehabilita- tion and return to training. Modes of exercise that theoretically provide fluid film conditions rather than boundary conditions appear prefer- ential. Studies of damaged cartilage (osteoarthrosis) and lubrication mechanisms suggest clinicians should be particularly attuned to the theo- ries that: 1 ) the manner (type) of loading is more important than the actual load. 2) the potential for wear increases when the joint goes from static loading to high loads, 3) syno- vial fluid volume and lymphatic clear- ance increase with exercise, 4) fric- tion is inversely related to the predicted thickness of the fluid film,

Following injury and periods of immobility,

cartilage is less stiff and less capable of

tolerating high loads.

5) cartilage deteriorates in the un- loaded state, and 6) repetitive impul- sive loading (eg., typing, using pneu- matic drill, skipping) is detrimental to both articular cartilage and s u b chondral bone.

Insufficient data exist to scientifi- cally prescribe ideal exercise pro- grams to ensure a fluid film mecha- nism and avoid cartilage damage. However, prudence suggests that in the presence of joint pathology, trau- ma, and/or effusion, passive motion should be accompanied by gentle traction to avoid compression of soft- ened cartilage and overstretching of vulnerable inflamed soft tissues, such as ligaments and tendons. Pendular- type movements provide a fluid film and minimal or no wear. Static load- ing, such as isometric exercise if

JOSPT Volume 28 Number 4 October 1998

used, should be very brief (a few sec- onds) as these conditions favor a mixed model of lubrication with boundary friction predominating. High loading should either be very brief or avoided. It is possibly least harmful following low-load, high ve- locity motion since these conditions favor the presence of a fluid film re- gion. If high loads are sustained for less than 0.5 seconds, a squeeze film mechanism should ensure a fluid film exists. If resistance is applied, in any form, the individual should work through the range, and holding peri- ods at end range (ie., knee exten- sion) should be very brief. Controlled eccentric loading creates large com- pressive, ie., loading forces; these should be avoided where cartilage damage may exist.

Research suggests that cartilage has such a low permeability that any fluid flow out of cartilage to contrib Ute to lubrication ofjoint surfaces is negligible. Therefore, clinicians can- not assume that load-bearing exer- cises will provide an adequate fluid film by promoting fluid efflux from cartilage. However, load bearing will promote fluid influx into cartilage and facilitate cartilage nutrition. It is anticipated that future research will elucidate this area for which minimal "hard" evidence currently exists and it is hoped that physical therapists will contribute to this knowledge with both basic and clinical studies. Areas to be addressed include: Are rehabili- tative outcomes improved following a regime of graduated alternating load and no load exercises? Is cartilage integrity over the long term en- hanced by restriction of sustained loaded conditions for a defined time period following insult? Are rehabili- tative outcomes (time to return to sport/employment/length of compet- itive sports career) improved follow- ing a regime that excludes isometric holds and controlled eccentric exer- cise in the initial 3 months postin- jury?

Investigators and therapists will need to follow for years, in fact, de-

cades, patients who have pathology or therapy that either directly or indi- rectly jeopardizes cartilage health if the influence of different regimes is to be properly assessed; long-term longitudinal studies are required.

CONCLUSION

Loading appears essential to maintain healthy articular cartilage although the physiological limits of normal (safe) loading remain as yet undefined. Conditions of no load, reduced load, and excessive load are less well tolerated and often are detri- mental to the health of cartilage. In therapy and training prescriptions, therapists should consider the rate, quantity, and purpose of loading.

JOSPT

REFERENCES 1. Aaron RK, Ciomber DM, lolly G:

Modulation of chondrogenesis and chondrocyte differentiation by pulsed electromagnetic fields. Trans Orthop Res Soc 12:272, 1987 (abstract)

2. Aaron RK, Plaas AK: Stimulation of proteoglycan synthesis in articular chondrocyte cultures by a pulsed electromagnetic field. Trans Orthop Res Soc 12:273, 1987 (abstract)

3. Akeson WH, Eichelberger L, Roma M: Biochemical studies of articular carti- lage. 11. Values following denervation of an extremity. ] Boneloint Surg 40A: 153- 162, 1 958

4. Akeson WH, Woo SL-Y, Amiel D, Coutts RD, Daniel D: The connective tissue response to immobility: Bio- chemical changes in periarticular tis- sue of the immobilized rabbit knee. Clin Orthop 93:356-362, 1973

5. Altman R, Asch E, Bloch D, Bole G, Borenstein D, Brandt K, Christy W, Cooke TD, Greenwald R, Hochberg M, Howell D, Kaplan D, Koopman W, Longley S, Mankin H, McShane Dl, Medsger T lr, Meenan R, Mikkelsen W, Moskowitz R, Murphy W, Roths- child B, Segal M, Sokoloff L, Wolfe F: Development of criteria for the clas- sification and reporting of osteoarthri- tis. Arthritis Rheum 29: 1039-1 04 1, 1986

6. Amiel D, Coutts RD, Harwood FL, Ishizue KK, Kleiner jB: The chondro- genesis of perichondrial grafts for re- pair of full thickness articular cartilage

defects in a rabbit model: A one-year post-operative assessment. Connect Tissue Res l8:27-39, 1988

7. Andersson S, Nilsson B, Hessel T, Sarate M, Noren A, Stevens-Anders- son A, Rydholm D: Degenerative joint disease in ballet dancers. Clin Orthop 238:233-236, 1989

8. Armstrong PF, Brighton CT, Star AM: Capacitatively coupled electrical stim- ulation of bovine growth plate chon- drocytes grown in pellet form. / Or- thop Res 6:265-27 1, 1 988

9. Arokoski 1: Altered Properties of Artic- ular Cartilage and Subchondral Bone After Long-Term Joint Loading. An Ex- perimental Study on Young Beagles, pp 1-62. Kuopio, Finland: Kuopio University Publications D. Medical Sciences 105, 1 996

10. Arokoski 1, lurvelin 1, Kiviranta I, Tammi M, Helminen HI: Softening of the lateral condylar articular cartilage in the canine knee joint after long dis- tance (up to 40 km/day) running train- ing lasting one year. Int ] Sports Med l5:254 -260, 1 994

11. Arokoski 1, Kiviranta I, lurvelin 1, Tammi M, Helminen HI: Long-dis- tance running causes site-dependent decrease of cartilage glycosaminogly- can content in the knee joint of beagle dogs. Arthritis Rheum 36: 145 1- 1459, 1993

12. Baker B, Becker OR, Spadaro 1: A study of electrochemical enhance- ment of articular cartilage repair. Clin Orthop 1 O2:25 1-267, 1 974

13. Baxendale RH, Ferrel WR, Wood L: Intra-articular pressures during active and passive moment of normal and distended human knee joints. ] Physiol 369: 179, 1983 (abstract)

14. Bennett GA, Waine H, Bauer W: Changes in the Knee joint at Various Ages, With Particular Reference to the Nature and Development of Degener- ative loint Disease, New York: The Commonwealth Fund, 1942

15. Brandt KD, Mankin H): Pathogenesis of osteoarthritis. In: Kelley WN, Harris ED, Ruddy S, Sledge CB (eds), Text- book of Rheumatology (4th Ed) (Vol- ume 2), pp 1355-1 373. Philadelphia: W.B. Saunders Company, 1993

16. Brandt KD, Mankin HI: Osteoarthritis and polychondritis. In: Kelley WN, Harris ED, Ruddy S, Sledge CB (eds), Textbook of Rheumatology (4th Ed) (Volume 2), pp 1355- 1373. Philadel- phia: W. B. Saunders Company, 1993

17. Brighton CT, Unger AS, Stambough ]L: In vitro growth of bovine articular car- tilage chondrocytes in various capaci-

Volume 28 Number 4 October 1998 JOSPT

C L I N I C A L C O M M E N T A R Y - .. - - -. . . -

tatively coupled electric fields. ] Or- thop Res 2: 75-22, 1984

18. Bruggemann GP: Biomechanics in gymnastics. Med Sport Sci Exerc 45: 142- 176, 1987

19. Bruns I, Volkmer M, Luessenhop S: Pressure distribution in the knee joint. Influence of flexion with and without ligament dissection. Arch Orthop Trauma Surg 1 l3:2O4 -209, 1 994

20. Buckwalter ]A, Cruess R: Healing of musculoskeletal tissues. In: Rock- wood CA, Green DP, Bucholz RW (eds), Fractures in Adults (3rd Ed), pp 18 1-222. Philadelphia: I. B. Lippincott Company, 1991

2 1. BuckwalterlS, Mow VC: Articular car- tilage repair in osteoarthritis. In: How- ell DS, Mankin HI, Moskowitz RW (eds), Osteoarthritis: Diagnosis and Management (2nd Ed), pp 7 1 - 107. Philadelphia: W.B. Saunders Compa- ny, 1990

22. Buckwalter ]A, Rosenberg LC, Hun- ziker EB: Articular cartilage: Compo- sition structure, response to injury and methods of facilitating repair. In: Ew- ing ]W (ed), Articular Cartilage and Knee loint Function: Basic Science and Arthroscopy, pp 19-56. New York: Raven Press, 1990

23. Bullough PG, Walker PS: The distribu- tion of load through the knee joint and its possible significance to the ob- served patterns of articular cartilage breakdown. Bull Hosp loint Dis Or- thop lnst 37: 1 10-123, 1977

24. Candolin T, Videman T: Effects of run- ning on the scanning electron micro- scopic appearance of joint surfaces of rabbits. Scand J Rheumat0160 (Suppl): 3 9, 1 986 (abstract)

25. Caterson B, Lowther DA: Changes in the metabolism of the proteoglycans from sheep articular cartilage in re- sponse to mechanical forces. Biochem Biophys A ~ t a 540:4 12-422, 1978

26. Chatham WW, Swaim R, Frohsin Hlr , Heck L W, Miller El, Blackburn WDlr : Degradation of human articular carti- lage by neutrophils in synovial fluid. Arthritis Rheum 36(1):5 1-58, 1993

27. Collins DH: The Pathology of Articu- lar and Spinal Diseases, London: Ar- nold, 1949

28. Coutts RD, Woo SL, Amiel D, von Schroeder HP, Kwan MK: Rib peri- chondrial autografts in full-thickness articular cartilage defects in rabbits. Clin Orthop 275:263-273, 1992

29. Crelin ES, Southwick WO: Changes induced by sustained pressure in the knee joint articular cartilage of adult rabbits. Anat Rec l49( l ): 1 13- 133, 1964

30. Eichelberger L, MileslS, Roma M: The histochemical characterization of ar- ticular cartilages of poliomyelitis pa- tients. / Lab Clin Med 40:284-296, 1 952

3 1. Ekholm R, Norbeck 8: On the rela- tionship between articular changes and function. Acta Orthop Scand 2 1: 81-98, 1951

32. Elliott ]P, Smith RL, Block CA: Time- varying magnetic fields: Effects of ori- entation on chondrocyte proliferation. 1 Orthop Res 6:259-264, 1988

33. Ely LW, Mensor MC: Studies on the immobilization of the normal joints. Surg Gynecol Obstet 37:2 12-2 15, 1933

34. Engelmark VE: Functionally induced changes in articular cartilage. In: Evans FG (ed), Biomechanical Studies of the Musculoskeletal System, pp 3-19. Springfield, IL: Charles C. Thomas, 1961

35. Enneking WF, Horowitz M: The intra- articular effects of immobilization of the human knee. 1 Bone loint Surg 54Ar973-985, 1972

36. Evans CH, Brown TD: Role ofphysical and mechanical agents in degrading the matrix. In: Woessner IF, Howell DS (eds), Joint Cartilage Degradation, p 199. New York: Marcel Dekker Inc., 1993

37. Evans EB, Eggers GWN, Butler ]K, Blumel 1: Experimental immobiliza- tion and remobilization of rat knee joints. ] Boneloint Surg 42A:737-758, 1960

38. Folkman J, Brean H: Angiogenesis and inflammation. In: Gallin 11, Goldstein IM, Snyderman R (eds), Inflammation: Basic Principles and Clinical Corre- lates (2nd Ed), pp 821-840. New York: Raven Press, 1992

39. Fox RI, Kang H: Structure and function of synoviocytes. In: McCarty Dl, Koopman WJ (eds), Arthritis and Al- lied Conditions (12th Ed) (Volume l), pp 263-278. Philadelphia: Lea & Febiger, 1993

40. Geborek P, Wollheim FA: Synovial fluid. In: Wright V, Radin E (eds), Me- chanics of Human joints, pp 109- 736. New York: Marcel Dekker, 1993

4 1. Gritzka TL, Fry LR, Cheesman RL, La Vigne A: Deterioration of articular cartilage caused by continuous com- pression in a moving rabbit joint. / Bone loint Surg 55A: 1698-1 720, 1973

42. Hall MC: Articular changes in the knee of the adult rat after prolonged immobilization in extension. Clin Or- thop 34:184-195, 1964

43. Hall MC: Cartilage changes after ex- perimental immobilization of the knee joint of the young rat. ] Boneloint Surg 45A:36-44, 1963

44. Harris ED ]r: Pathogenesis of rheuma- toid arthritis. In: Kelley WN, Harris ED, Ruddy S, Sledge CB (eds), Text- book of Rheumatology (2nd Ed) (Vol- ume I), pp 886-915. Philadelphia: W.B. Saunders Company, 1985

45. Havdrup T, Telhag H: Scattered mito- sis in adult joint cartilage after partial chondrectomy. Acta Orthop Scand 48:424-429, 1978

46. Helminen H], lurvelin 1, Kuusela T, Heikkila R, Kiviranta I, Tammi M: Ef- fects of immobilization for 6 weeks on rabbit knee articular surfaces as as- sessed by the semi-quantitative ste- reomicroscopic method. Acta Anat 1 15:327-335, 1983

47. Helminen HI, Kiviranta I, Saamanen A-M, Jurvelin], Arokoski A, Oettmeier R, Abendroth K, Roth A], Tamrni MI: Effect of motion and load on articular cartilage in animal models. In: Kuett- ner K, Schleyerbach R, Peyron /G, Hascall VC (eds), Articular Cartilage and Osteoarthritis, pp 501-51 0. New York: Raven Press, 1992

48. Holmdahl DE, Ingelmark BE: Der bau des gelenkknorpels unter ver- schiedenen funktionellen verhaltnis- sen. Experimentelle untersuchung an wachsenden kaninchen. Acta Anat 6: 120-126, 1948

49. lnsall 1: Patella pain. ] Boneloint Surg 64A: 147- 152, 1 982

50. lurvelin 1, Helminen H], Lauritsalo S, Kiviranta I, Saamanen A-M, Pauk- konen K, Tammi M: Influences ofjoint immobilization and running exercise on articular cartilage surfaces of young rabbits. Acta Anat 122:62-68, 1985

5 1. Jurvelin J, Kiviranta I, Saamanen A-M, Tammi M, Helminen HJ: Indentation stiffness of young canine knee articu- lar cartilagdnfluence of strenuous joint loading. / Biornech 23: 1239-1246, 1990

52. lurvelin], Kuusela T, Heikkila R, Pelt- tari A, Kiviranta I, Tammi M, Helmi- nen HI: Investigation of articular car- tilage surface morphology with a semiquantitative scanning electron microscopic method. Acta Anat 116: 302-3 1 1, 1 983

53. Kellgren IH, Lawrence IS: Radiologi- cal assessment of osteoarthrosis. Ann Rheum Dis 16:494-501, 1957

54. Kiss I, Morocz I, Herczeg L: Localiza- tion and frequency of degenerative changes from the knee joint: Evalua-

JOSPT Volume 28 Number 4 October 1998

tion of 200 necropsies. Acta Morphol Hung 32: 155- 163, 1 984

55. Kiviranta I, Tammi M, Jurvelin J, Sad- manen A-M, Helminen HI: Moderate running exercise augments glycosami- noglycans and thickness of articular cartilage in the knee joint of young beagle dogs. J Orthop Res 6: 188- 1 95, 1988

56. Kujala UM, Kettunen J, Paananen H, Aalto T, Battie MC, lmpivaara 0, Vite- man T, Sarna S: Knee osteoarthritis in former runners, soccer players, weight lifters, and shooters. Arthritis Rheum 38:539-546, 1995

57. Krause W-D: Mikroskopische untersu- chungen am gelenkknorpel extrem funktionell belasteter mause-ein bei- trag zur aetiopathogenese degenera- tiver gelenkveranderungen. Unpub- lished dissertation, University of Koln, Koln, Germany, 1969

58. Landells JW: The reactions of injured human articular cartilage. J Bone Joint Surg 39B:548-562, 1 957

59. Lane NE, Michel B, Bjorkengren A: The risk of osteoarthritis with running and aging: A 5-year longitudinal study. J Rheumatol20:46 1-468, 1993

60. Lanier RR: The effects of exercise on the knee-joints of inbred mice. Anat Rec 94:3 1 1-32 1, 1946

6 1. Larsson T, Aspden RM, Heinegard D: Effects of mechanical loading on car- tilage matrix biosynthesis in vitro. Ma- trix 1 1 (6):388-394, 1991

62. Lippiello L, Chakkalakal D, Connolly JF: Pulsing direct current-induced re- pair of articular defects. J Orthop Res 8(2):266-275, 1990

63. Lohmander LS: Osteoarthritis: Man, models and molecular markers. Intro- duction. In: Maroudas A, Kuettner K (eds), Methods in Cartilage Research, p 337. New York: Academic Press, 1990

64. Mankin HI, Dorfman H, Lippiello L, Zarins A: Biochemical and meta- bolic abnormalities in articular cartilage from osteo-arthritic human hips. J Bone Joint Surg 53A:523-537, 1971

65. Mankin HI, Radin EL: Structure and function of joints. In: McCarty Dl, Koopman WJ (eds), Arthritis and Al- lied Conditions (12th Ed) (Volume l), pp 181-1 97. Philadelphia: Lea & Fe- biger, 1993

66. Maroudas A: Biophysical chemistry of cartilaginous tissues with special ref- erence to solute and fluid transport. Biorheology 12:233-248, 1975

67. Martel-Pelletier J, Pelletier J-P, Clou- tier J-M, Howell DS, Ghandur- Mna ymneh L, Woessner IF Jr: Neutral

proteases capable of proteoglycan di- gesting activity in osteoarthritic and normal human articular cartilage. Ar- thritis Rheum 27:305-3 12, 1984

68. McCarty Dl: Clinical picture of rheu- matoid arthritis. In: McCarty Dl, Koopman WJ (eds), Arthritis and Al- lied Conditions (12th Ed) (Volume l), pp 781-810. Philadelphia: Lea & Fe- biger, 1993

69. Meachim G, Roy S: Surface ultrastruc- ture of mature adult human articular cartilage. J Bone Joint Surg 5 1B:529- 539, 1969

70. Mitrovic D, Borda-lriate 0, Naveau B, Stankovic A, Uzan H, Quintero M, Ryckewaert A: Results of autopsy ex- amination of knee cartilage of 120 pa- tients dying in the hospital. ll. The femoro-tibia1 joint. Rev Rhum Ma1 Os- teoartic 56505-5 10, 1989

71. Moran ME, Kim HK, Salter RB: Biolog- ical resurfacing of full-thickness de- fects in patella articular cartilage of the rabbit. Investigation of autogenous periosteal grafts subjected to continu- ous passive motion. J Bone Joint Surg 74B:659-667, 1992

72. O'Driscoll SW, Salter RB: The induc- tion of neochondrogenesis in free in- tra-articular periosteal autografts un- der the influence of continuous passive motion: An experimental study in the rabbit. J Bone Joint Surg 66A:1248-1257, 1984

73. Oegema TR: Selection and classifica- tion of articular cartilage. In: Marou- das A, Kuettner K (eds), Methods in Cartilage Research, pp 7-8. New York: Academic Press, 1990

74. Okihana H, Shimomura Y: Effect of direct current on cultured growth car- tilage cells in vitro. J Orthop Res 6:690-694, 1988

75. Palmoski MI, Brandt KD: Effects of static and cyclic compressive load- ing on articular cartilage plugs in vitro. Arthritis Rheum 27:675- 68 1, 1984

76. Palmoski MJ, Colyer RA, Brandt KD: Joint motion in the absence of normal loading does not maintain normal ar- ticular cartilage. Arthritis Rheum 23: 325-334, 1980

77. Parkkinen JJ, Ikonen J, Lammi MJ, Laa- konen J, Tammi M, Helminen HI: Ef- fects of cyclic hydrostatic pressure on proteoglycan synthesis in cultured chondrocytes and articular cartilage explants. Arch Biochem Biophys 300: 458-465, 1993

78. Paukkonen K, Helminen HI: Decrease of proteoglycan granule number but increase in the size in articular carti- lage of young rabbits after physical

exercise and immobilization by splint- ing. Anat Rec 2 19:45-52, 1987

79. Paukkonen K, Selkanaho K, Jurvelin 1, Kiviranta I, Helminen HI: Cells and nuclei of articular cartilage chondro- cytes in young rabbits enlarged after non-strenuous physical exercise. J Anat 142: 73-20, 7985

80. Poole AR: Cartilage in health and dis- ease. In: McCarty Dl, Koopman WJ (eds), Arthritis and Allied Conditions (12th Ed) (Volume 11, pp 279-333. Philadelphia: Lea & Febiger, 1993

8 1. Prockop Dl, Williams CJ, Vandenburg P: Collagen in normal and diseased connective tissue. In: McCarty Dl, Koopman WJ (eds), Arthritis and Al- lied Conditions (12th Ed) (Volume l), pp 2 13-227. Philadelphia: Lea & Fe- biger, 1993

82. Radin EL: Mechanically inducedperi- articular and neuromuscular prob- lems. In: Wright V, Radin EL (eds), Mechanics of Human Joints, pp 355- 370. New York: Marcel Dekker, Inc., 1993

83. Radin EL, Ehrlich MGM, Chernack R, Abernethy PI, Paul IL, Rose RM: Effect of repetitive impulse loading on the knee joints of rabbits. Clin Orthop 131:288-293, 1978

84. R e p RU, Finlay JB: Survival of articular cartilage after controlled impact. J Bone Joint Surg 59A: 1068-1 076, 1977

85. Rodosky MW, Fu FH: Induction of sy- novial inflammation by matrix mole- cules, implant particles, and chemical agents. In: Leadbetter WB, Buckwalter ]A, Gordon SL (eds), Sports-Induced Inflammation, pp 357-38 1. Park Ridge, IL: American Academy of Or- thopaedic Surgeons, 1990

86. Sadf J: Effects of exercise on adult ar- ticular cartilage. Acta Orthop Scand 7 (Suppl7): 1-86, 1950

87. Saamanen A-M, Tammi M, Kiviranta 1, Helminen HI: Running exercise as a modulator of proteoglycan matrix in the articular cartilage of young rabbits. Int J Spom Med 9: 127- 132, 1 988

88. Sadmanen A-M, Tammi M, Kiviranta I, Jurvelin J, Helminen HI: Levels of chondroitin-6-sulfate and nonaggre- gating proteoglycans at articular carti- lage contact sites in the knees of young dogs subjected to moderate running exercise. Arthritis Rheum 32: 1282-1292, 7989

89. Salter B: Continuous Passive Motion (CPM): A Biological Concept for the Healing and the Generation of Artic- ular Cartilage, Ligaments, and Ten- dons: From its Origination to Research to Clinical Applications, Baltimore: Williams & Wilkins, 1993

Volume 28 Number 4 October 1998 JOSPT

90. Salter RB, Field P: The effects of con- tinuous compression on living articu- lar cartilage. ] Bone Joint Surg 42A: 31-49, 1960

91. Schultz R], Krishnamurthy S, Thelmo W, Rodriquez JE, Harvey G: Effects of varying intensities of laser energy on articular cartilage: A preliminary study. Laser Surg Med 5577-588, 1 985

92. Seedholm BB, Takeda T, Tsubuku, Wright V: Mechanical factors and patellofemoral osteoarthrosis. Ann Rheum Dis 38:307-3 16, 1979

93. Shortkroff S, Barone L, Hsu H-P, Wrenn C, Gagne T, Chi T, Breinan H, Minas T, Sledge CB, Tubo R, Spector M: Healing of chondral and osteo- chondral defects in a canine model: The role of cultured chondrocytes in regeneration of articular cartilage. Biomaterials 17: 147- 154, 1 996

94. Simkin PA: Synovial physiology. In: McCarty Dl, Koopman W] (eds), Ar- thritis and Allied Conditions (12th Ed) (Volume I), pp 199-212. Philadel- phia: Lea & Febiger, 1993

95. Simon MR, Holmes KR, Olsen AM: Effects of simulated increases in body weight on the growth of limb bones in hypophysectomized rats. Acta Anat 121:l-6, 1985

96. Sledge CB: Biology of the joint. In: Kelley WN, Harris ED, Ruddy S, Sledge CB (eds), Textbook of Rheuma- tology (4th Ed) (Volume 2), pp 1-21. Philadelphia: W. B. Saunders Compa- ny, 1993

97. Sokoloff L: The Biology of Degenera- tiveloint Disease, Chicago: University of Chicago Press, 1 969

98. Sood SC: A study of the effects of ex- perimental immobilization on rabbit articular cartilage. ] Anat 108:497-507, 1971

99. Stockwell RA: Structure and function of the chondrocyte under mechanical stress. In: Helminen HI, Kiviranta I, Sadmdnen A-M, Tammi M, Pauk- konen K, lurvelin ] (eds), Joint Load- ing, pp 126- 148. Bristol: Wright, 1987

100. Tammi M, Kiviranta I, Peltonen L, lurvelin 1, Helminen HI: Effects of joint loading on articular cartilage collagen metabolism: Assay of procol- lagen, prolyl 4-hydrozylase and galactosylhydn~xyIsyI, glycmyltansfe- nase. Connect Tissue Res 17: 199-206, 1 988

101. Tammi M, Paukkonen K, Kiviranta I, lurvelin 1, Saamanen A-M, Helminen HI: loint loading-induced alterations in articular cartilage. In: Helminen HI, Kiviranta I, Saamanen A-M, Tammi M, Paukkonen K, lurvelin ] (eds), joint Loading, pp 126-148. Bristol: Wright, 1987

102. Tammi M, Sadmanen A-M, jauhiainen A, Malminen 0, Kiviranta I, Helminen H: Proteoglycan alterations in rabbit knee articular cartilage following physical exercise and immobilization. Connect Tissue Res 16: 163- 175, 1 983

103. Telhag H: Mitosis of chondrocytes in experimental osteoarthritis in rabbits. Clin Orthop 86:224 -229, 1972

104. Thompson RC, Bassett CAL: Histolog- ical observations on experimentally induced degeneration of articular car- tilage. ] Bone Joint Surg 52A:435- 443, 1970

105. Threlkeld A], Smith SD: Unilateral hind paw amputation causes bilateral articular cartilage remodelling of the rat hip joint. Anat Rec 221:576-583, 1988

106. Tomlinson N, Maroudas A: The effect of cyclic and continuous compression on the penetration of large molecules into articular cartilage. ] Bone Joint Surg 626325 1, 1980 (abstract)

107. Trias A: Effect of persistent pressure on articular cartilage. An experimental study. I Boneloint Surg 43 B:376 -386, 1961

108. Trueta 1: Osteo-arthritis: An approach to surgical treatment. Lancet 1:585- 589, 1956

109. van Kampen GPI, van de Stadt lR: Cartilage and chondrocyte responses to mechanical loading in vitro. In: Helminen HI, Kiviranta I, Sadmdnen

A-M, Tammi M, Paukkonen K, lurve- lin ] (eds), loint Loading, pp 1 12-125. Bristol: Wright, 1987

1 10. Vasan N: Effects of physical stress on the synthesis and degradation of car- tilage matrix. Connect Tissue Res 12: 49 -58, 1 983

1 1 1. Veldhuizen ]P, Bourret LA, Rodan GA: In vitro studies of the effect of intermittent compressive forces on cartilage cell proliferation. 1 Cell Physiol 98:299-306, 1979

112. Verzar F: Lectures on Experimental Gerontology, Springfield, IL: Charles C. Thomas, 1963

1 13. Videman T, Eronen 1: Effects of tread- mill running on glycosaminoglycans in articular cartilage of rabbits. Int ] Sports Med 5:320-324, 1984

114. Videman T, Eronen I, Candolin T: Ef- fects of motion load changes on ten- don tissues and articular cartilage. A biochemical and scanning electron microscopic study on rabbits. Scand l Work Environ Health (Suppl 5):56- 67, 1979

1 15. Viidik A: Functional properties of col- lagenous tissues. Int Rev Connect Tis- sue Res 6: 127-2 15, 1 973

1 16. Walker IM: Exercise and its influence on aging in rat knee joints. ] Orthop Sports Phys Ther 8:3 10-3 1 9, 1986

1 17. Wandhofer A, Bally GV, Kauffmann G, Karduck A, Richter H-G: Compar- ative morphological investigation on the rabbit's auricle after exposure to CO, and Nd:YAElaser radiation. Arch Otorhinolaryngol2 7 7:487-49 1, 1977

118. Williams IM, Moran M, Thonar El- MA, Salter RB: Continuous passive motion stimulates repair of rabbit knee articular cartilage after matrix proteoglycan loss. Clin Orthop 304: 252-262, 1 994

1 19. Zvaifler N]: Etiology and pathogenesis of rheumatoid arthritis. In: McCarty Dl, Koopman Wl (eds), Arthritis and Allied Conditions (12th Ed) (Volume I), pp 723-736. Philadelphia: Lea & Febiger, 1993

JOSF'T Volume 28 Number 4. October 1998