Metabolic changes in rabbit articular cartilage due to inflammation.

The effect of inflammation on the articular cartilage of rabbit knee joints were studied. The inflammation was induced by intraarticular injections of corton oil or rabbit peritoneal leukocyte lysates. An increase in the activities of various lysosomal enzymes was observed in the synovial fluid as well as in the cartilage of the inflamed joints. Loss of proteoglycans, increased rate of degradation of collagen and proteoglycans, and increased rate of their synthesis were evident in the treated cartilage. The rate of uptake of 3H-thymidine was also increased. A significant change was observed in the type of collagen synthesized by these explants in vitro. In addition to synthesizing their characteristics Type II collagen, the cartilage explants from the treated joints synthesized Type I collagen.     

Metabolic changes in rabbit articular cartilage due to inflammation

The effects of inflammation on the articular cartilage of rabbit knee joints were studied. The inflammation was induced by intraarticular injections of croton oil or rabbit peritoneal leukocyte lysates. An increase in the activities of various lysosomal enzymes was observed in the synovial fluid as well as in the cartilage of the inflamed joints. Loss of proteoglycans, increased rate of degradation of collagen and proteoglycans, and increased rate of their synthesis were evident in the treated cartilage. The rate of uptake of ³H-thymidine was also increased. A significant change was observed in the type of collagen synthesized by these explants in vitro. In addition to synthesizing their characteristic Type II collagen, the cartilage explants from the treated joints synthesized Type I collagen.     

Effects of tiaprofenic acid (Surgam) on cartilage proteoglycans in the rabbit joint immobilisation model.

A well established model of arthritis induced in rabbit knee joints by immobilisation in full extension for 30 days was used to evaluate the in vivo effects of 2.5, 5.0, and 10.0 mg/kg body weight of tiaprofenic acid on articular cartilage proteoglycans. The drug was given subcutaneously every 24 hours during the entire immobilisation period. Immobilised animals not treated with drugs and normal animals were used as controls. In the non-drug treated immobilised animals articular cartilage showed evidence of surface damage accompanied by synovial hypertrophy and effusion. Proteoglycan concentrations were reduced in cartilages of these joints and the incorporation of 35SO2-4 into macromolecular proteoglycans was higher than in cartilages of non-immobilised controls. Gel filtration chromatographic studies of both resident and 35S labelled proteoglycans isolated from immobilised joint cartilage showed reduced aggregation and the presence of degraded proteoglycan subunit species. Whereas the administration of 10.0 mg/kg tiaprofenic acid every 24 hours to immobilised animals exacerbated the degradation and loss of proteoglycans from joint cartilages, 5.0 mg/kg tiaprofenic acid provided some protection of these macromolecules, as shown by the concentrations and extractability of proteoglycans from cartilages, which were comparable with those from non-immobilised controls. A high incorporation of 35S into proteoglycans was demonstrated, together with reduced catabolism of proteoglycans, indicating preservation of chondrocyte anabolic activity. At a tiaprofenic acid dose of 2.5 mg/kg, however, no beneficial effects on cartilage proteoglycans could be shown.     

Prevention of cartilage destruction with intraarticular osteoclastogenesis inhibitory factor/osteoprotegerin in a murine model of osteoarthritis

OBJECTIVE: To investigate the effect of osteoclastogenesis inhibitory factor/osteoprotegerin (OPG) on chondrocytes in the development of osteoarthritis (OA) in vivo. METHODS: To determine the role of endogenous OPG in the progression of OA, OA was surgically induced in OPG+/- mice and their wild-type (WT) littermates. To determine the role of exogenous OPG, knee joints of C57BL/6J mice with surgically induced OA were injected intraarticularly with recombinant human OPG (rHuOPG) or vehicle 5 times a week. All mice were euthanized 4 weeks after OA induction; joints were harvested and evaluated immunohistochemically. RESULTS: Although OA changes were induced in both WT and OPG+/- mice, the degenerative changes in the articular cartilage were significantly enhanced in OPG+/- mice. In C57BL/6J mice with surgically induced OA, intraarticular OPG administration protected the articular cartilage from the progression of OA. The Mankin and cartilage destruction scores in OPG-treated animals were approximately 50% of those seen in the control group. Furthermore, OPG administration significantly protected articular cartilage thickness. Findings of the TUNEL assay indicated that rHuOPG prevented chondrocyte apoptosis in joints with surgically induced OA. Results of immunostaining indicated that OPG protein was detected in the synovium and in resident chondrocytes at higher levels in the OPG-treated group than in the control group. CONCLUSION: These data indicate that endogenous OPG had a protective effect against the cartilage destruction that occurs during OA progression. Furthermore, direct administration of rHuOPG to articular chondrocytes prevented cartilage destruction in an experimental murine model of OA via prevention of chondrocyte apoptosis.     

Mechanical effects of the intraarticular administration of high molecular weight hyaluronic acid plus phospholipid on synovial joint lubrication and prevention of articular cartilage degeneration in experimental osteoarthritis

OBJECTIVE: To examine in vivo the effects of a mixture of high molecular weight hyaluronic acid (HA) plus phospholipids on joint lubrication and articular cartilage degeneration. METHODS: Experimental osteoarthritis (OA) of the right knee was induced by anterior cruciate and medial collateral ligament transection in 40 rabbits. The animals were subjected to 8 consecutive weekly intraarticular administrations of high molecular weight HA (the HA200 group), conventional molecular weight HA (the HA80 group), or high molecular weight HA plus L-delta dipalmitoyl phosphatidylcholine liposomes (the PHA group) and were killed 1 week after the final injection. The remaining transected right knees (the OA group) and randomly selected nontransected contralateral left knees (the control group) were collected simultaneously. Each group (n = 10) was divided into 2 equal subgroups, one of which was evaluated histologically while the other was subjected to a lubricating ability test using a pendulum friction tester. RESULTS: The injected knees had a tendency to demonstrate less damage to the articular cartilage compared with the OA group, and the histologic findings in all groups except for the PHA group differed significantly from the control group. There was a significant difference in the mean +/- SD friction coefficient between the control group (0.0100 +/- 0.00300) and the OA (0.0206 +/- 0.00649), HA200 (0.0190 +/- 0.00427), and HA80 (0.0177 +/- 0.00712) groups (P < 0.05 for each comparison), but not between the control group and the PHA group (0.0150 +/- 0.00330) (P = 0.15). CONCLUSION: To our knowledge, this is the first in vivo study to examine whether intraarticular injections of phospholipids influence joint lubrication by acting as a boundary lubricant, thus protecting articular cartilage from degenerative changes.     

Water-soluble C60 fullerene prevents degeneration of articular cartilage in osteoarthritis via down-regulation of chondrocyte catabolic activity and inhibition of cartilage degeneration during disease development

OBJECTIVE: Studies have shown the roles of oxidative stress in the pathogenesis of osteoarthritis (OA) and induction of chondrocyte senescence during OA progression. The aim of this study was to examine the potential of a strong free-radical scavenger, water-soluble fullerene (C60), as a protective agent against catabolic stress-induced degeneration of articular cartilage in OA, both in vitro and in vivo. METHODS: In the presence or absence of C60 (100 microM), human chondrocytes were incubated with interleukin-1beta (10 ng/ml) or H2O2 (100 microM), and chondrocyte activity was analyzed. An animal model of OA was produced in rabbits by resection of the medial meniscus and medial collateral ligament. Rabbits were divided into 5 subgroups: sham operation or treatment with C60 at 0.1 microM, 1 microM, 10 microM, or 40 microM. The left knee joint was injected intraarticularly with water-soluble C60 (2 ml), while, as a control, the right knee joint received 50% polyethylene glycol (2 ml), once weekly for 4 weeks or 8 weeks. Knee bone and cartilage tissue were prepared for histologic analysis. In addition, in the OA rabbit model, the effect of C60 (10 microM) on degeneration of articular cartilage was compared with that of sodium hyaluronate (HA) (5 mg/ml). RESULTS: C60 (100 microM) inhibited the catabolic stress-induced production of matrix-degrading enzymes (matrix metalloproteinases 1, 3, and 13), down-regulation of matrix production, and apoptosis and premature senescence in human chondrocytes in vitro. In rabbits with OA, treatment with water-soluble C60 significantly reduced articular cartilage degeneration, whereas control knee joints showed progression of cartilage degeneration with time. This inhibitory effect was dose dependent, and was superior to that of HA. Combined treatment with C60 and HA yielded a significant reduction in cartilage degeneration compared with either treatment alone. CONCLUSION: The results indicate that C60 fullerene is a potential therapeutic agent for the protection of articular cartilage against progression of OA.     

Effect of salicylate on proteoglycan metabolism in normal canine articular cartilage in vitro.

In osteoarthritis, diminished aggregation of articular cartilage proteoglycans affects tissue biomechanics. Since salicylates are commonly employed in treatment of osteoarthritis, we examined the effect of sodium salicylate on proteoglycan metabolism and aggregation in normal canine articular cartilage. At salicylate concentrations of 10(-3)M, 5 X 10(-3)M and 10(-2)M, net proteoglycan synthesis in normal canine articular cartilage was 73%, 42% and 16% respectively, of control levels. Catabolism of glycosaminoglycans in the presence of 10(-3)M salicylate (which corresponds to a serum salicylate level of 20-25 mg %) was the same as that in control cartilage, while higher concentrations of the drug increased the rate of degradation. The hydrodynamic size of newly synthesized proteoglycan aggregates and of disaggregated proteoglycans was unaffected by sodium salicylate.     

Articular responses to purified cartilage proteoglycans

To examine whether proteoglycans (PGs) liberated from cartilage might contribute to articular changes in arthritis, cartilage PGs were injected intraarticularly into rabbit knee joints. Twice-weekly injections of PG (2.5 mg) provoked synovial hypertrophy, synovitis, erosion of the articulating surfaces, and loss of metachromasia of the articular cartilage. These changes were accompanied by a marked elevation in the production of neutral collagenase and gelatinase by both synoviocytes and chondrocytes. The synoviocytes of experimental knee joints also produced factor(s), possibly related to interleukin-1, which provoked the activation of chondrocytes. Our data are consistent with the idea that free PG fragments mediate some of the pathophysiologic changes that occur in arthritic joints. This property may be particularly important in osteoarthritis.     

Methylprednisolone acetate induced release of cartilage proteoglycans: determination by high performance liquid chromatography.

A high performance liquid chromatography (HPLC) procedure suitable for the simultaneous determination of the molecular size and concentration of macromolecular hyaluronate and proteoglycans in synovial fluid has been developed. Irrigation of the equine tarsocrural joint with 20 ml physiological saline (PSS) caused a mild inflammation with an increase of proteoglycans in the synovial fluid over the baseline arthrocentesis control sample. Proteoglycan and hyaluronate in the synovial fluid did not interact to form hyaluronate-proteoglycan aggregates, but separated as distinct chromatographic peaks. This suggests that the cartilage derived proteoglycans in synovial fluid in the inflamed joint have been proteolytically cleaved from the non-covalent aggregates containing link protein and hyaluronate. Hyaluronidase digestion completely abolished the hyaluronate peak without affecting the proteoglycans. This seems to indicate that proteoglycan in synovial fluid is unable to interact with hyaluronate in synovial fluid to form cartilage type aggregates. Proteolytic degradation and the time dependent release into the synovial fluid of such digested proteoglycan also resulted from the intra-articular injection of methylprednisolone acetate into normal tarsocrural joints and joints irrigated with PSS. These proteoglycans were insensitive to hyaluronidase but may consist of a protein moiety with attached glycosaminoglycans, as suggested by their sensitivity to proteinase and keratanase/chondroitinase digestion. These observations with cartilage treated with methylprednisolone acetate and mildly stimulated articular cartilage are inconsistent with earlier work on osteoarthritic and rheumatoid articular cartilage and have interesting implications for the pathogenesis and for the therapeutic action of intraarticular corticosteroids. A rapid HPLC procedure applicable to unprocessed small volume samples of synovial fluid gives information simultaneously on hyaluronate and proteoglycan in synovial fluid which is not attainable with immunoradiometric or isotope tracer techniques. It therefore appears to be useful for the analysis of cartilage turnover and destruction in health and disease.     

Adenovirus-mediated gene transfer of insulin-like growth factor 1 stimulates proteoglycan synthesis in rabbit joints

OBJECTIVE: To examine the effect of insulin-like growth factor 1 (IGF-1) on the regulation of cartilage synthesis and other articular events in vivo. METHODS: A first-generation adenoviral vector expressing human IGF-1 (AdIGF-1) from the cytomegalovirus promoter was constructed. Particles of AdIGF-1 (5 x 10(9)) were injected through the patellar tendon into normal rabbit knee joints and rabbit knee joints with antigen-induced arthritis (AIA), with the same dose of a control adenoviral vector injected into the contralateral knees. Lavage fluids were obtained from rabbit knee joints on days 3 and 7 postinjection and used for analysis of IGF-1 expression, white blood cell infiltration, and cartilage breakdown. Cartilage chips from rabbit joints were used for assay of new proteoglycan synthesis, and tissues also were harvested from the dissected knees for histologic study. RESULTS: Intraarticular injection of AdIGF-1 resulted in a mean of 180.6 ng/ml of IGF-1 expression in the lavage fluid from rabbit joints. IGF-1 expression stimulated new proteoglycan synthesis in both naive and AIA rabbit knees, but had no significant chondroprotective or antiinflammatory effects. Histologic analysis showed that elevated levels of IGF-1 expression in both normal and arthritic knees had no adverse pathologic effects on synovium or adjacent muscles. CONCLUSION: Gene transfer of IGF-1 into rabbit knee joints promotes proteoglycan synthesis without significantly affecting inflammation or cartilage breakdown. In addition, no adverse effects following intraarticular IGF-1 gene delivery were observed. Thus, local gene transfer of IGF-1 to joints could serve as a therapeutic strategy to stimulate new matrix synthesis in both rheumatoid arthritis and osteoarthritis.     

Celecoxib has a positive effect on the overall metabolism of hyaluronan and proteoglycans in human osteoarthritic cartilage

OBJECTIVE: To assess the effects of celecoxib, a cyclooxygenase (COX-2) selective inhibitor, on the metabolism of hyaluronan (HA) and proteoglycans (PG) in human cartilage explants with midrange severity of osteoarthritis (OA). Results were compared with those of diclofenac, a non-selective COX inhibitor. METHODS: Cartilage specimens (OA grade 4-8 on Mankin's scale) were pulsed with 3H -glucosamine and chased in the absence or presence of 1-10 micro g/ml of celecoxib or diclofenac. After papain digestion, the labeled chondroitin sulfate and HA molecules were purified by anion-exchange chromatography. RESULTS: Diclofenac did not affect the metabolic balance of PG and HA whereas, in a relatively dose-dependent manner, celecoxib increased the synthesis of HA and PG; celecoxib also reduced the net loss of labeled HA and PG molecules from cartilage explants. CONCLUSION: In short term in vitro cultures, celecoxib has a favorable effect on the overall metabolism of PG and HA. It is therefore unlikely that this drug would have a detrimental effect on articular cartilage during longterm administration. Further, celecoxib might help counteract the depletion of HA seen in OA cartilage.     

Carrageenin-induced arthritis.1. The effect of intraarticular carrageenin on the chemical composition of articular cartilage

A single intraarticular injection of carrageenin into rabbit knee joints initiated a localized synovial inflammatory response. This response was accompanied by a 20% loss of proteoglycan from the articular cartilage within 24 hours and by a further 30–60% loss within 5–7 days. The chondrocytes replaced the lost proteoglycan within 42 days. More than two injections caused only a further small decrease in proteoglycan content; the cartilage was then unable to replace the lost proteoglycan. The absence of recovery coincided with the appearance of erosion of the cartilage surface.     

Intra-articular Hyaluronic Acid following Knee Immobilisation for 6 Weeks in Rabbits

Thirty-two mature female New Zealand White rabbits were immobilised in an aluminium splint on one knee, using the contralateral non-treated knee as control. After 6 weeks the splints were removed and the rabbits allowed unrestricted movement. On a random basis, 16 rabbits were given an intra-articular injection of 5 mg in 0.5 ml of hyaluronic acid (HA) in the knee of the immobilised hindlimb at weekly intervals for 6 weeks, starting 1 week after the joint had been remobilised, for a total of six injections. In the other 16 rabbits no further intervention was conducted. At the end of the experiment the rabbits were killed, the area of degenerated joint surface of the distal femur, and water and proteoglycan content were measured, and the articular cartilage stained with haematoxylin and eosin and safranin O. Remobilisation without HA administration resulted in a significantly larger degenerated joint surface area. By the end of the experiment both remobilisation and remobilisation and intra-articular HA injections had produced a greater but non-significant water cartilage content compared to the control side. The average cartilage glycosaminoglycan content of the remobilisation and intra-articular HA injection group was significantly greater than in the remobilisation group. In conclusion, in the rabbits with one knee immobilised for 6 weeks, 6 weeks of remobilisation alone are not sufficient to recover from the moderate articular surface changes produced, and the intra-articular administration of HA may produce a morphologically and biochemically more normal cartilage. More extensive animal and human studies should be performed before the routine use of intra-articular administration of HA following musculoskeletal injuries that required immobilisation can be recommended. Received: 25 January 2000 / Accepted: 11 August 2000     

Suppression of hyaluronic acid synthesis in synovial organ cultures by corticosteroid suspensions

The effects of sparingly soluble corticosteroid suspensions prepared for intraarticular therapy, of their vehicles, and of hydrocortisone on synovial hyaluronic acid (HA) synthesis were compared in organ cultures of normal canine villous synovium. Both hydrocortisone and the corticosteroid suspensions suppressed HA synthesis, as did 2 vehicle components, polysorbate 80 and myristyl-gamma-picrolinium chloride. Cultures of synovium from joints of dogs which, 1 day previously, had been injected with methylprednisolone acetate suspension synthesized less HA than did control cultures from noninjected joints of the same animals. The results indicate that suppression of HA synthesis is a mechanism by which these drugs can act to reduce joint HA content and promote resolution of synovial effusions.     

Metalloproteinases and cartilage proteoglycan depletion in chronic arthritis

Chronic monarticular arthritis can be induced in ovalbumin-sensitized rabbits by intraarticular injection of ovalbumin (antigen-induced arthritis) or in naive rabbits by injecting hyaluronic acid mixed with the polycation poly-D-lysine (polycation-induced arthritis). Both models show some points of similarity, including joint swelling, the presence of inflammatory leukocytes and the inflammatory mediator prostaglandin E₂, and the kinetics of cartilage proteoglycan loss. However, the assessment of the capacity of synovial lining and articular cartilage to synthesize and secrete neutral metalloproteinases reveals a difference between these models. We found that articular cartilage from the inflamed joints of rabbits with antigen-induced arthritis did not synthesize neutral metalloproteinases, although the synovial lining did. In contrast, both the synovial lining and the articular cartilage from the inflamed joints of rabbits with polycation-induced arthritis synthesized neutral metalloproteinases. These findings suggest that in inflammatory synovitis, different mechanisms can operate to produce damage to the matrix of articular cartilage.     

Effect of hyaluronan on chondrocyte apoptosis and nitric oxide production in experimentally induced osteoarthritis

OBJECTIVE: Nitric oxide (NO) plays an important role in cartilage degeneration, and NO donors induce chondrocyte apoptosis. This study evaluated the effect of intraarticular injections of hyaluronan (HA) on chondrocyte apoptosis and NO production using an experimental osteoarthritis (OA) model. METHODS: Thirty-six New Zealand white rabbits underwent unilateral anterior cruciate ligament transection (ACLT), and were divided into 3 groups. Four weeks after ACLT, the HA group started intraarticular HA injections once a week for 5 weeks; the vehicle group started to receive the carrier of HA; the no injection group received no treatment. All ACLT knees were harvested at Week 9 and evaluated for OA severity. Culture supernatants of the cartilage were analyzed for nitrite concentration. Cartilage sections were analyzed by TUNEL for apoptosis and by immunohistochemistry for nitrotyrosine. RESULTS: OA in the HA group was less severe than the other 2 groups. The number of apoptotic chondrocytes was significantly lower in the HA group. HA injection did not influence NO production in cartilage. CONCLUSION: HA protects against chondrocyte apoptosis during the development of OA, while it may not have definite effects on NO production in the joints. These inhibitory effects of HA on chondrocyte apoptosis may play a role in its mechanism of action in chondroprotection.     

Antigen-induced experimental septic arthritis in rabbits after intraarticular injection of Staphylococcus aureus

With the Dumonde-Glynn model of antigen-induced arthritis, a rabbit model was developed to examine the histopathologic differences between normal and arthritic joints in the same animal infected by intraarticular injections of Staphylococcus aureus. Microscopic examination of whole joint sections and a quantitative histopathologic scale were used to compare changes in all the articular components of 17 normal and 17 arthritic joints infected for less than two weeks. The histological changes were more severe in infected arthritic joints than in infected normal joints (mean +/- SD total histology score, 13.8 +/- 2.4 and 9.3 +/- 4.0, respectively; P less than .001). In infected arthritic joints, subsynovial abscesses extended into subchondral bone via the pannus of chronic synovitis at articular margins and intraarticular attachments of cruciate ligaments, rather than by initial cartilage destruction and direct extension into subchondral bone.     

EFFECT OF SALICYLATE ON PROTEOGLYCAN METABOLISM IN NORMAL CANINE ARTICULAR CARTILAGE IN VITRO

In osteoarthritis, diminished aggregation of articular cartilage proteoglycans affects tissue biomechanics. Since salicylates are commonly employed in treatment of osteoarthritis, we examined the effect of sodium salicylate on proteoglycan metabolism and aggregation in normal canine articular cartilage. At salicylate concentrations of 10⁻³ M, 5 x 10⁻³ M and 10⁻² M, net proteoglycan synthesis in normal canine articular cartilage was 73%, 42%, and 16%, respectively, of control levels. Catabolism of glycosaminoglycans in the presence of 10⁻³ M salicylate (which corresponds to a serum salicylate level of 20–25 mg %) was the same as that in control cartilage, while higher concentrations of the drug increased the rate of degradation. The hydrodynamic size of newly synthesized proteoglycan aggregates and of disaggregated proteoglycans was unaffected by sodium salicylate.     

Aspirin aggravates the degeneration of canine joint cartilage caused by immobilization

The effect of aspirin on the degeneration of knee cartilage caused by immobilization was examined. If dogs were fed aspirin daily (serum salicylate = 20–25 mg/dl) for 6 weeks while one hind limb was immobilized in a cast, the decreases in uronic acid content and net proteoglycan synthesis in cartilage from the immobilized knee were significantly greater than the decreases in cartilage from immobilized knees of dogs that had not received aspirin (P<0.01). Furthermore, neither aspirin administration nor immobilization alone affected the extractability of proteoglycans from the cartilage. However, in organ cultures of cartilage from the immobilized knee of dogs fed aspirin, the proportion of the total ³⁵S-proteoglycans present in the culture medium was nearly twice that from cultures of cartilage of the contralateral knee. Also, more than twice as many of the total tissue proteoglycans (uronic acid) were extractable with 0.4M guanidinium chloride, a nondissociating solvent (P<0.01). Regardless of whether the dogs received aspirin, the in vitro interaction of proteoglycans with hyaluronic acid from cartilage of the immobilized knee was diminished, apparently due to an abnormality in the hyaluronate-binding region of the core protein. Although these results indicate that aspirin had an adverse effect in vivo on articular cartilage of immobilized joints, aspirin administration did not preclude reversal of all of the above changes if the dog was allowed to walk about in a pen for 3 weeks after cast removal. If, however, the dog was run daily on a treadmill for 3 weeks after cast removal, the decrease in uronic acid content in cartilage from the immobilized knee persisted and was more profound if the animal had received aspirin than if it had not (P<0.01).     

Fibronectin on the surface of articular cartilage in rheumatoid arthritis

The presence of fibronectin on the surface of articular cartilage in rheumatoid arthritis (RA) was investigated. Cartilage samples were stained by the immunoperoxidase method using anti-human fibronectin antibody, and observed under light and electron microscopy. Fibronectin was present on the articular surface in 7 of 8 RA patients. The degree of staining varied greatly among the patients. Five of 8 patients were positive for fibronectin in 50% or more of the cartilage areas studied. In total, fibronectin was observed in RA. Fibronectin was not observed in cartilage samples of osteoarthritic joints or joints which were not diseased but had undergone trauma. Ultrastructurally, it was observed to be associated with collagen fibrils and amorphous substance in the matrix. The fibronectin-negative surface of the rheumatoid cartilage was usually thick ultrastructurally, compared with the fibronectin-positive surface, and the staining for fibronectin roughly correlated with decreased proteoglycans on the surface. The presence of fibronectin in the matrix appeared to be revealed by partial degradation of proteoglycans with proteolytic enzymes in the synovial fluid, as well as by the deposition of fibronectin onto the surface of rheumatoid cartilage. Fibronectin on the articular surface may play an important role in promoting pannus extension onto the articular surface in RA.