Long-term effects of hyaluronan on experimental osteoarthritis in the rabbit knee

Osteoarthritis and Cartilage (1998) 6, 1–9 7 1998 Osteoarthritis Research Society

1063–4584/98/010001 + 09 $12.00/0

Long-term effects of hyaluronan on experimental osteoarthritis in the rabbit knee By Choji Shimizu, Makoto Yoshioka, Richard D. Coutts, Frederick L. Harwood, T. Kubo*, Y. Hirasawa* and David Amiel Department of Orthopaedics, University of California San Diego, La Jolla CA; Malcolm and Dorothy Coutts Institute for Joint Reconstruction and Research, San Diego CA, U.S.A. and *Department of Orthopaedic Surgery, Kyoto Prefectural University of Medicine, Kyoto, Japan

Summary Objective: Long-term assessment of the effect of hyaluronan (HA) on the articular cartilage and synovium in an animal knee joint during the development of osteoarthritis (OA). Design: Sixty mature New Zealand white rabbits underwent unilateral anterior cruciate ligament transection (ACLT) and were divided into two groups. Group 1 (SA) received intra-articular injections of 0.3 ml hyaluronan (HA) (i.e., ARTZ, MW: 8 × 105) beginning 4 weeks after ACLT, once a week for 5 weeks. Group 2 (SV) received injections of the vehicle (phosphate buffered saline) in the same fashion as with the SA group. The contralateral nonoperated knee served as control. All animals were killed 21 weeks after surgery and their knee joints evaluated by gross morphologic, histologic, histomorphometric and biochemical analyses. Results: Gross morphological inspection indicated that the femoral condyles from the knees injected with vehicle suffered more severe cartilage damage than cartilage from the knees injected with HA. Furthermore, two out of three histomorphometric parameters measured in the HA-treated cartilage (i.e., cartilage thickness and cartilage area which were not statistically different than control) provided evidence showing a protective effect of HA on the femoral condyles following ACLT. Biochemical analysis showed articular cartilage remaining on the femoral condyles following ACLT to have similar characteristics to contralateral controls. However, DNA concentration in the synovium from the ACLT knees of the vehicle-treated animals was greater than contralateral control, while this parameter was not statistically different than contralateral control in the HA treated animals. Conclusions: These results demonstrate a protective effect of HA on preservation of the articulating surface of the femoral condyle following ACLT up to 21 weeks post-surgery. Key words: Hyaluronan, Osteoarthritis, Cartilage, Joints.

tration of hyaluronan (HA). Several studies have explored the relationship between HA and synovial fluid in normal and osteoarthritic joints [4–6], while a variety of clinical studies have reported on the relative effectiveness of treatment of osteoarthritis (OA) with HA of differing molecular weights [7–10]. Although little is known of HA’s in vivo mechanisms in the suppression of the clinical symptoms of OA, in vitro studies have shown that HA has the ability to modulate a variety of cellular functions as well as suppress the activities of specific pro-inflammatory mediators [11–13]. Recently several in vivo animal studies have reported on the early term effectiveness of HA treatment on experimentally induced osteoarthritis [14–18]. In a previous report from this laboratory, for instance, it was shown that in the near term, i.e. 9 weeks following ACL transection

Introduction Knee instability following a complete tear of the anterior cruciate ligament (ACL) often induces osteoarthritis with accompanying degradation of the articular cartilage matrix [1–3]. Non-surgical therapeutic treatment for such a condition has included, among other treatments, the adminis-

Received 25 March 1997; accepted 17 July 1997. The authors wish to acknowledge the financial support of the Malcolm & Dorothy Coutts Institute for Joint Reconstruction and Research, the Seikagaku Corporation and NIH grants AR28467 and AG07996. The authors have no commercial interest in the production or sale of the hyaluronan provided by the Seikagaku Corporation. Please address all correspondence to: David Amiel, Ph.D. Professor of Orthopaedics, Univ. Calif., San Diego, 9500 Gilman Dr Dept 0630, La Jolla CA 92093-0630. Phone: 619 534 2676 Fax: 619 534 5304.

1

2

Shimizu et al.: Longer term effect of HA

(ACLT), intra-articular injections of HA [molecular weight (MW) = 8 × 105 ] were effective in maintaining specific histomorphometric and biochemical parameters of the articular cartilage matrix [19]. Because of the reported short half-life of HA [21, 22], however, a question naturally arose regarding the effectiveness of HA treatment during development of osteoarthritis over the long-term. Therefore, the purpose of the present study was to carry out a controlled study on the effectiveness of HA treatment on the suppression of osteoarthritis over an extended period of time following joint destabilization. This study assesses the longer term (21 weeks post-injury) effects of HA on histologic, histomorphometric and biochemical parameters of the cartilage matrix following ACLT.

lazine (2 mg/kg). ACLT knees were then shaved and disinfected with Betadine solution. A 27-gauge needle on a tuberculin syringe was inserted beneath the patella in the region of the patellofemoral joint and 0.3 ml HA or vehicle were injected. Intra-articular injections were performed once a week for 5 weeks. All animals were killed 21 weeks following surgery by intracardiac injection of T-61 euthanasia solution (Fig. 1).

Materials and Methods

Histological preparation

experimental design

In accordance with the results of a previous study [3] in which the medial femoral condyles showed the most advanced changes, only medial femoral condyles were selected for histological preparation and assessment. Tissue blocks were fixed in 10% neutral buffered formalin with cetylpyridinium chloride (CPC) for 7 days and decalcified with EDTA [22, 23]. After decalcification (confirmed by X-ray), the femoral condyles were cut along the sagittal plane and the medial condyles were embedded in paraffin. Five micron sections were cut with a Reichert–Jung microtome and stained with safranin O/fast green. The synovium from the ACLT and contralateral knees of each group was also processed, embedded in paraffin and stained with hematoxylin and eosin (H & E) for cellular assessment by image analysis.

Sixty mature New Zealand white (NZW) rabbits underwent unilateral ACLT of the left knees. The contralateral nonoperated right knees served as controls. All rabbits were divided into two groups. Group 1 (SA) received 0.3 ml intra-articular injections of HA into the left knee 4 weeks after ACLT, once a week for 5 weeks. Group 2 (SV) was injected in the same manner with the vehicle (carrier of HA: sterile phosphate buffered saline). All animals were killed 21 weeks following surgery and were then divided into three subgroups for assessment of knee joints by gross morphologic, histologic/histomorphometric and biochemical analyses.

assessment Gross morphology Gross morphological changes to the femoral condyles were assessed as described in a previous study [19] following the application of india ink.

surgery and treatment of animals Histomorphometry All rabbits were anesthetized by intramuscular injection of ketamine (100 mg/kg) and xylazine (8 mg/kg). Operative knees were shaved and disinfected with Betadine solution. A medial parapatellar incision was made on the skin and a medial arthrotomy was performed. The patella was dislocated laterally and the ACL was transected. The joint was then irrigated with sterile saline and closed. The capsule was closed with a running suture of 4-0 prolene and the skin with a running mattress suture supplemented with interrupted sutures of 3-0 nylon. Postoperatively the animals were permitted cage (60 × 60 × 40 cm) activity. The animals were closely monitored for infections and other complications. Four weeks after surgery those rabbits receiving intra-articular injections of HA or vehicle were anesthetized with ketamine (30 mg/kg) and xy-

The histological sections were visualized and assessed using a color image analysis system (Oncor instrument systems, U.S.A.) with microscope (Nikon Microphot, Japan) and high-resolution color video camera (Hitachi HV-C10, Japan). Video images were captured by a highdefinition video processing board in the image analysis system computer (Everex 486/25, U.S.A.). SA group

4 weeks

Surgery

SV group

4 weeks

ARTZ injection

12 weeks

Once a week × 5 weeks

Sacrifice

Vehicle injection

12 weeks

21 weeks

Fig. 1. Surgery and treatment protocol: time scale. ( OA development; (,) treatment; (q) observation.

)

Osteoarthritis and Cartilage Vol. 6 No. 1

Images were viewed on a high-resolution color monitor (Sony Trinitron, Japan). Customized image analysis software measured the following geometric parameters: (1) thickness, (2) area, (3) surface roughness, and (4) thickness of synovial lining cell layer. The geometric parameters of the cartilage were measured on a condylar weight-bearing area of 7 mm length at ×32 magnification. The thickness of cartilage from the surface to the tide mark was calculated from the mean of five measurements made perpendicular to the surface of each section at five equally spaced points along the 7 mm length. The area of the cartilage present between the monitor screen edges (7 mm) was also calculated. The thickness and area of cartilage were also computed using a mathematical algorithm and employing the coordinates of the articular cartilage and tidemark. Calculation of the cartilage surface roughness is based on the deviations from an idealized smooth surface [19] and is expressed as the root mean square (RMS) surface roughness. Inflammation/cellular proliferation of synovial tissue was assessed by measuring the thickness of the synovial lining cell layer. This thickness was determined by averaging the measured thickness at 100 different locations along the length of the histological specimen.

biochemistry Water content Immediately after death the articular cartilage was dissected from the femoral condyles of the ACLT and contralateral control knees, placed into preweighed vials and weighed again to an accuracy of 0.1 mg. The samples were then lyophilized and reweighed for determination of water content. Non-reducible collagen cross-links (pyridinoline) concentration The concentration of the non-reducible collagen cross-link hydroxypyridinium contained in the articular cartilage was determined by reversephase high-performance liquid chromatography (HPLC) [24]. Five milligrams (dry) of tissue were hydrolyzed in 6 n HCl overnight at 108°C. Following desiccation, the hydrolysate was redissolved in 1% heptafluorobutyric acid and the cross-link hydroxypyridinium was isolated by HPLC on a reversed-phase column (ODS, C-18; Beckman Instruments U.S.A.). Elution was by a

3

gradient of acetonitrile (12–18%) in 0.01 m heptafluorobutyric acid. Monitoring and quantitation of hydroxypyridinium was achieved by in-line fluorescence (ex. = 297 nm, em. = 390 nm). The standard employed was pyridoxamine, which has identical spectral characteristics with hydroxypyridinium. Results are expressed as equivalent moles of pyridoxamine per mole of collagen. Hexosamine concentration Total hexosamine content in the tissue was determined by colorimetry [24]. Three to 5 mg of lyophilized cartilage were hydrolyzed in 1 ml of 6 n HCl for 5 h at 100°C. The acid hydrolysates were desiccated and hexosamine quantitation was achieved by a modified Elson–Morgan reaction. The results are expressed as micrograms of amino sugar per milligram of dry tissue. DNA concentration in synovial tissue Synovial tissue cellularity was assessed by quantitating the tissue concentration of DNA according to the procedure of Bonting and Jones as modified by Amiel et al. [24]. Seven to 10 mg H2O washed and lyophilized tissue were solubilized by incubation for 2 h in 3.0 ml of 1 n NaOH at 65°C. Duplicate 1.0 ml aliquots were reacted with 0.04% indole–HCl reagent and mixed with chloroform to remove interfering substances. The aqueous phase containing the DNA was removed and the absorbance was read at 490 nm. Standards of calf thymus DNA were also determined for comparison. Results are expressed as microgram DNA per milligram of dry tissue. statistical analysis All data are presented as mean 2 standard deviation and were subjected to statistical analysis using analysis of variance (ANOVA) with paired data sets, with a level of significance set at a = 0.05. Results gross morphology Femoral condyle Figures 2(a) and (b) illustrate visually the result of ACLT on the femoral condyles from the SA and SV groups, respectively. Osteophytes on the medial condyles were observed to be generally more prominent with a corresponding increase in severity of cartilage damage. Gross morphological grading of cartilage damage was performed on the medial femoral condyles and the results are illustrated in Fig. 3. Various grades of OA were

4

Shimizu et al.: Longer term effect of HA

changes except for one condyle which had grade 2 fibrillation. Synovium Thickening of the synovium relative to contralateral control was observed in all of the ACLT knees; there were no obvious differences in synovial hypertrophy between the SA and SV groups. qualitative histological evaluation Femoral condyle Histological evidence for cartilage degeneration (surface roughness and loss of safranin/O staining) was observed in the ACLT knees of both SA and SV groups. However, there appeared to be a rougher cartilage surface and less uniform safranin/O staining in the condyles from the SV group than in the condyles from the SA group. The contralateral condyles demonstrated a mostly normal histological appearance. Representative histological assessment is illustrated in Fig. 4a–c. Synovium Hyperplasia of the synovial lining cell layer was observed in the ACLT knees from both the SA and SV groups, with the SV synovia demonstrating slightly greater hyperplasia than the SA synovia (Fig. 5). histomorphometry—quantitative analysis Cartilage thickness and area Fig. 2: Gross morphology of the femoral condyles. (a) ACLT knee: grade 3 OA specimen from the SA group. Overt fibrillation is seen on the medial condyle (white arrows). (b) ACLT knee: grade 4c OA specimen from the SV group. Full-thickness ulceration is seen on the medial condyle (closed white arrow) along with overt fibrillation on the lateral condyle (open arrows). (c) Specimen of contralateral control condyle. Smooth cartilage surface shows normal appearance.

Statistically significant decreases (relative to contralateral controls) in cartilage thickness and Contralateral control SA n = 10

10

9

observed in the ACLT knees from both groups; however the severity of damage was generally greater in the SV group compared with the SA group. Whereas in the SA group only one condyle out of 10 evaluated showed grade 4c cartilage surface damage, five condyles in the SV group demonstrated grade 4c damage. The contralateral condyles gave no evidence of osteoarthritic

ACLT

1

SV n = 10

Grade 1

Grade 4a

Grade 2

Grade 4b

Grade 3

Grade 4c

1

2

2

4

2

1

1 1 1

5

Fig. 3. Gross morphological assessment of the femoral condyles from SA and SV groups (N = 10 each). Note that the extent and grade of cartilage damage in the condyles from the ACLT knees in SA group was less severe than in the SV group.

Osteoarthritis and Cartilage Vol. 6 No. 1

5

Cartilage surface roughness In the SV group, surface roughness of the femoral condyle cartilage from the ACLT knees was statistically greater than contralateral control while this parameter was not significantly different from control in the SA group. These results are shown in Fig. 6(c).

Fig. 4. Histological evaluation (Safranin O/fast green) of the femoralcondyles. (a) ACLT specimen from the SA group (original magnification × 30). Irregularity of the cartilage surface is seen in accordance with the loss of Safranin O stain. (b) ACLT specimen from the SV group (original magnification × 30). Cartilage thickness is decreased with severe fibrillation. Ulceration of the cartilage exposing the smooth surface of the subchondral bone is seen. (c) Contralateral control specimen (original magnification × 30). Cartilage matrix was consistently well-stained with Safranin O.

area were observed in the ACLT knees from the SV group. Cartilage thickness and area were not statistically different from controls in the SA group. These data are illustrated in Fig. 6(a) and (b).

Fig. 5. Histological evaluation (H&E) of synovium. (a) ACLT knee specimen from the SA group (original magnification × 93). Hyperplasia of synovial lining cell layer was seen in part (white arrow). (b) ACLT knee specimen from the SV group (original magnification × 93). Hyperplasia of synovial lining cells is marked (white arrow). (c) Contralateral control knee specimen (original magnification × 93). Proliferation of the synovial lining cell layer is minimal.

6

Shimizu et al.: Longer term effect of HA 0.4

3

(a) *

* Area (mm2)

0.3 Thickness (mm)

(b)

0.2

2

1

0.1

0

0

20

(c)

RMS roughness

0.06

*

0.04

0.02

0

SA

SA

SV

SV

Thickness (microns)

0.08

SA

15

SV

(d)

*

*

10

5

0

SA

SV

Fig. 6. Histomorphometric analysis of the medial femoral condyles from HA and vehicle treated knees. All data expressed as mean 2 standard deviation. (a) Cartilage thickness: N = 10; *P Q 0.05. (b) Cartilage area: N = 10; *P Q 0.05. (c) Cartilage roughness: N = 10; *P Q 0.05. (d) Thickness of synovial lining cell layer: N = 10; *P Q 0.05. (Q) ACLT; ( ) Control.

Thickness of the synovial lining cell layer

Hexosamine concentration

Thickness of the ACLT synovial lining cell layer in the SA and SV groups was significantly greater than their respective contralateral controls. This parameter is shown in Fig. 6(d).

The hexosamine concentrations of the femoral cartilage in the ACLT knees were not statistically different than their respective control values. This was true for both SA and SV groups [Fig. 7(c)].

biochemistry

DNA concentration in synovium

Water content

In the SA group no statistically significant difference in DNA content was observed in the synovium from the ACLT knees relative to contralateral control. However, in the SV group, the DNA concentration in the synovium was significantly higher than its contralateral control. These results are illustrated in Fig. 7(d).

The water content of the ACLT cartilage from the femoral condyle in the SA and SV groups was similar to their respective controls, i.e., near-normal values with no statistically significant differences observed [Fig. 7(a)]. Pyridinoline concentration In the SA and SV groups no significant differences in pyridinoline concentration were observed in the femoral condyle cartilage from the ACLT knees relative to their respective contralateral controls [Fig. 7(b)].

Discussion Joint destabilization secondary to a complete tear of the ACL is known to cause a breakdown of the articulating surfaces in those joints with the resulting loss of joint function [1–3]. In a previous

Osteoarthritis and Cartilage Vol. 6 No. 1

study we showed that in the short-term (following ACLT), HA had beneficial effects in preserving assessments of the articular cartilage and synovium were carried out. Since these assessments were made immediately after the end of HA treatment, it was important to address whether HA, in the ACLT model, might have longer-term beneficial effects on the articular cartilage and synovium of the injured joints. In the present study we report that HA, despite its reported short half-life [20, 21], slowed progression of OA in the articular cartilage and synovium 12 weeks after cessation of treatment. Morphological grading of the femoral condyles revealed that the extent and severity of cartilage damage was less in the HA treated than in the vehicle treated knees. Furthermore, histomorphometric analysis showed that cartilage thickness, area and roughness in the vehicle treated knees were all statistically inferior to their contralateral controls, while these par-

ameters in the HA treated knees were not significantly different than controls. The results of biochemical analysis of the articular cartilage were less conclusive. There was an apparent discrepancy between the hexosamine concentration [Fig. 7(c)] data and the safranin O staining shown in Fig. 4. The examples of safranin O stained sections include a ‘good’ HA-treated specimen and a ‘poor’ vehicle-treated specimen. These results varied somewhat from specimen to specimen, although morphological grading showed that overall the HA-treated cartilage suffered less severe changes than the vehicle-treated cartilage. The hexosamine data, on the other hand, is a mean from 10 specimens in each group, and these results showed that, on average, no difference between groups could be established. Also, no differences between ACLT cartilage and contralateral control cartilage were seen in either treatment group for the other biochemical parameters. There was,

3

100

(b) Moles of pyridoxamine per mole of collagen

(a)

Percent water

75

50

25

SA

1

SA

SV

60

SV

20

(c)

µg DNA per mg dry tissue

(d)

45

µg Hexosamine per mg dry tissue

2

0

0

30

15

0

7

SA

SV

15

* 10

5

0

SA

SV

Fig. 7. Biochemical analysis of the medial femoral condyles from HA andvehicle treated knees. All data expressed as mean 2 standard deviation. (a) Water content in the articular cartilage. N = 10. (b) Pyridinoline concentration in the articular cartilage: N = 10. (c) Hexosamine concentration in the articular cartilage: N = 10. (d) DNA concentration in the synovium: N = 10; *P Q 0.05. (Q) ACLT; ( ) Control.

8

Shimizu et al.: Longer term effect of HA

however, an inherent problem in the biochemical evaluation since the analyses were made on the specific histomorphometric and biochemical parameters of the articular cartilage and synovium in the affected joints [19]. In that study the animals were killed immediately after HA treatment and histologic, histomorphometric and biochemical cartilage remaining following cartilage loss over the 21-week post-ACLT period. Thus, the biochemical characteristics of the ACLT cartilage in the two treatment groups might appear similar to their contralateral controls and to each other even while actual loss of cartilage surface was greater in one group compared with the other. Indeed, the morphological and histomorphometric data revealed greater degeneration of the vehicle treated ACLT cartilage than HA treated ACLT cartilage. In this long-term study the most prominent biochemical difference was that the DNA concentration in the ACLT synovium from the vehicletreated knee was statistically higher than its contralateral control (indicating synovial inflammation), while this parameter was similar to control in the HA treated synovium. This was consistent with the histological (H & E) observation that synovial hyperplasia appeared more pronounced in the SV group than in the SA group. From these data we can postulate an anti-inflammatory effect of HA on the rabbit knee joint 12 weeks after the end of treatment. This conclusion is supported by the findings of Yasui [25] and Aihara [26] who demonstrated a reduced level of prostaglandin E2 PGE2 and bradykinin in the synovial joint following HA injection. In terms of the direct effect of exogenous HA on the articular cartilage matrix components, Yasui et al. [6] showed in an in vitro study that the addition of HA stimulates TIMP-1 production in bovine articular chondrocytes. The feedback mechanism of exogenous HA in the synovial joint, as postulated by Iwata [27], strongly supports the concept of upregulation of the endogenous HA since the half-life of exogenously administered HA was shown to be no more than 22 h [20]. In summary, the morphologic, histomorphometric and biochemical results of this long-term study indicate that HA has a protective effect on the articular cartilage up to 21-weeks postACLT. To our knowledge this is the first study to establish such long-term in vivo beneficial effects of exogenous HA on the preservation of articular cartilage and synovium following joint destabilization in an animal model.

Acknowledgments We acknowledge Karen Bowden, Michael Amiel and Rob Healey for their technical assistance. The authors wish to acknowledge the financial support of the Malcolm & Dorothy Coutts Institute for Joint Reconstruction and Research, the Seikagaku Corporation and NIH grants AR28467 and AG07996.

References 1. Brandt KD, Braunstein EM, Jisco DM, O’Conner B, Heck D, Albrecht M. Anterior cruciate ligament transection in the dog: a bona fide model of osteoarthritis, not merely of cartilage injury and repair. J Rheumatol 1991;18:436–46. 2. Mcdevitt CA, Muir H. An experimental model of osteoarthritis: early morphological and biochemical changes. J Bone Joint Surg 1977;59B:24–35. 3. Yoshioka M, Coutts RD, Amiel D, Hacker SA. Characterization of a model of osteoarthritis in the rabbit knee. Osteoarthritis Cart 1996;4:87–98. 4. Balazs EA, Watson D, Duff IF, Roseman S. Hyaluronic acid in synovial fluid I: molecular parameters of hyaluronic acid in normal and arthritic human synovial fluids. Arthritis Rheum 1967;10:357–66. 5. Balazs EA. The physical properties of synovial fluid and the special role of hyaluronic acid. In: Helfert AJ, Ed. Disorders of the Knee. Lippincott, Philadelphia, 1974, pp. 61–74. 6. Yasui T, Akatsuka M, Tobetto K, et al. Effects of hyaluronan on the production of stromelysin and tissue inhibitor of metalloproteinase-1 (TIMP-1) in bovine articular chondrocyte. Biomed Res 1992;13:343–48. 7. Helfet AJ. Management of osteoarthritis of the knee joint. In: Helfet AJ, Ed. Disorders of the Knee. Lippincott, Philadelphia, 1974, pp. 175–94. 8. Namiki O, Toyoshima H, Morisaki N. Therapeutic effect of intra-articular injection of high molecular weight hyaluronic acid on osteoarthritis of the knee. Int J Clin Pharmacol Ther Toxicol 1982;20:501–7. 9. Bragantini A, Cassini M, De Bastiani G, Perbellini A. Controlled single-blind trial of intra-articularly injected hyaluronic acid (Hyalgan) in osteoarthritis of the knee. Clin Trials J 1987;24:330–40. 10. Dixon AJ, Jacoby RK, Berry H, Hamilton EBD. Clinical trial of intra-articular injection of sodium hyaluronate in patients with osteoarthritis of the knee. Curr Med Res Opinion 1988;11:205–13. 11. Hakansson L, Hallgren R, Venge P. Regulation of granulocyte function of hyaluronic acid in vitro and in vivo effects on phagocytes, locomotion and metabolism. J Clin Invest 1980;66:298–305. 12. Forrester JV, Balazs EA. Inhibition of phagocytosis by high molecular weight hyaluronate. Immunology 1980;40(3):435–46. 13. Tamoto K, Tada M, Shimada S, Nochi H, Mori Y. Effect of high-molecular-weight hyaluronans on the function of guinea pig polymorphonuclear leukocytes. Semin Arthritis Rheum 1993;22 (Suppl. 1):4–8. 14. Abatangelo G, Botti P, Bue D, Gei G, Samson JC, Cortivo R. Intra-articular sodium hyaluronate injections in the Pond-Nuki experimental model of

Osteoarthritis and Cartilage Vol. 6 No. 1

15.

16.

17.

18.

19.

20.

osteoarthritis in dogs. I. Biochemical results. Clin Orthop 1989;241:278–85. Kitoh Y, Katsuramaki T, Tanaka H, et al. Effect of SL-1010 (sodium hyaluronate with high molecular weight) on experimental osteoarthritis induced by intra-articularly applied papain in rabbits. Folia Pharmacol Japan 1992;100:67–76. Ghosh P, Read R, Armstrong S, Wilson D, Marshall R. The effect of intra-articular administration of hyaluronan in a model of early osteoarthritis in sheep I: gait analysis and radiological and morphological studies. Semin Arthritis Rheum 1993;22(Suppl. 1):18–30. Ghosh P, Read R, Numata Y, Smith S, Armstrong S, Wilson D. The effects of intraarticular administration of hyaluronan in a model of early osteoarthritis in sheep. Vol II: cartilage composition and proteoglycan metabolism. Semin Arthritis Rheum 1993;22(Suppl. 1):31–42. Yoshimi T, Kikuchi T, Obara T, et al. Effects of high-molecular-weight sodium hyaluronate on experimental osteoarthrosis induced by the resection of rabbit anterior cruciate ligament. Clin Orthop 1994;298:296–304. Yoshioka M, Shimizu C, Harwood FL, Coutts RD, Amiel D. The effect of hyaluronan during the development of osteoarthritis. Osteoarthritis Cart 1997;5(4):257–60. Fraser JRE, Kimpton WG, Piersionek BK, Cahill RNP. The kinetics of hyaluronan in normal and acutely inflamed synovial joints: Exploratory

21.

22.

23. 24.

25.

26.

27.

9

observations with experimental arthritis in sheep. Semin Arthritis Rheum 1993;22(Suppl. 1): 9–17. Laurent UBG, Fraser JRE, Angstrom-Laurent A, Reed RK, Dahl LB, Saurent TC. Catabolism of hyaluronan in the knee joint of the rabbit. Matrix 1992;12:130–36. Engfeldt B, Hjertquist S-O. The effect of various fixatives on the preservation of acid glycosaminoglycans in tissues. Pathol Microbiol Scand 1967;171:219–32. Jubb RB, Eggert FM. Staining of demineralized cartilage II. Histochem 1981;73:391–96. Amiel D, Kleiner JB, Roux RD, Harwood FL, Akeson WH. The phenomenon of ligamentization: Anterior cruciate ligament (ACL) reconstruction with autogenous patellar tendon (PT). J Orthop Res 1986;4(2):162–172. Yasui T, Akatsuka K, Tobbeto M, Hayaishi M, Ando T. The effect of hyaluronan on interleukin-1ainduced prostaglandin E2 production in human osteoarthritic synovial cells. Agents Actions 1992;37:155–56. Aihara S, Mukakami N, Ishii R, et al. Effects of sodium hyaluronate on the nociceptive response of rats with experimentally-induced arthritis. Folia Pharmacol Japan 1992; 100: 359–65. Iwata H. Pharmacologic and clinical aspects of intra-articular injection of hyaluronate. Clin Orthop 1993;289:285–91.