Modulation of cartilage destruction by select nonsteroidal antiinflammatory drugs.

257
MODULATION OF CARTILAGE DESTRUCTION BY
SELECT NONSTEROIDAL ANTIINFLAMMATORY
DRUGS
In Vitro Effect on the Synthesis and Activity of Catabolism-Inducing Cytokines
Produced by Osteoarthritic and Rheumatoid Synovial Tissue
JEROME H. HERMAN, ALAN M. APPEL, and EVELYN V. HESS
Non-enzymatic factors produced by synovial tissue can potentially mediate cartilage destruction by
inducing the synthesis and release of matrix-degrading
proteinases from chondrocytes. Pharmacologic control
of this process is of potential clinical relevance. The in
vitro effect of therapeutic concentrations of select
nonsteroidal antiinflammatory drugs on the synthesis
and activity of catabolism-inducing cytokines produced
by 6-day explant cultures of osteoarthritic and rheumatoid synovial tissue was studied. Piroxicam regularly
suppressed such factor synthesis by both types of tissue without significantly affecting total protein synthesis. This did not occur using sodium salicylate
or indomethacin in osteoarthritis tissue cultures and was
observed only occasionally in rheumatoid arthritis
cultures. None of the nonsteroidal antiinflammatory
drugs studied consistently blocked catabolism-inducing
activity in osteoarthritis tissue, whereas piroxicam more
consistently inhibited activity produced by rheumatoid
arthritis tissue. Results suggest that the catabolisminducing factors produced by the 2 tissue sources may
differ.
From the Division of Immunology, Department of Internal
Medicine, IJniversity of Cincinnati Medical Center, Cincinnati,
Ohio.
Supported in part by grants AM-16894 and AM-20615 from
the NIH, an Arthritis Foundation Clinical Research Center grant,
and a grant from Pfizer Laboratories.
Jerome H. Herman, MD: Professor of Medicine, Department of Internal Medicine; Alan M. Appel, PhD: Research Assistant
Professor, Department of Internal Medicine; Evelyn V. Hess, MD:
Professor of Medicine and Director, Division of Immunology,
Department of Internal Medicine.
Address reprint requests to Jerome H. Herman, MD,
Division of Immunology, ML 563, University of Cincinnati Medical
Center, 231 Bethesda Avenue, Cincinnati, OH 45267.
Submitted for publication February 24, 1986; accepted in
revised form September 8, 1986.
Arthritis and Rheumatism, Vol. 30, No. 3 (March 1987)
Although nonsteroidal antiinflammatory drugs
(NSAIDs) may effectively suppress articular inflammation, whether they are capable of modifying the
natural course of cartilage destruction in joint disease
is unknown. This possibility is suggested in experimental animal models of adjuvant arthritis (1-3), antigen-induced arthritis (4), type I1 collagen-induced
arthritis (5-7), and osteoarthritis (OA) (8,9) in which
the administration of select classes of NSAIDs has
been associated with diminished pannus formation,
cartilage erosion, and/or bone destruction.
Although a given type of arthritis may have
unique precipitating features, most forms of disease
probably share common pathway mechanisms which
induce and perpetuate cartilage destruction (10). One
of the earliest manifestations in the incipient stage of
cartilage involvement is a focal, or at times more
diffuse, depletion of matrix proteoglycans. This appears to be primarily enzyme-mediated; however, the
source of responsible proteinases remains unclear.
The significance of enzymes evolving from synovial
fluid must be questioned in that pathologic effusions
contain proteinase inhibitors (1 1,12). Enzymes released in the immediate proximity of cartilage are
probably of greater importance. They may originate in
granulation tissue pannus, or as a consequence of
phagocytosis or reversed endocytosis associated with
immune complex deposition in the superficial
collagenous layer of cartilage (13,14). Such sources of
enzymes, however, would only provide a partial explanation for matrix destruction in that proteoglycan
depletion may, at times, be observed in midzone and
deeper regions of cartilage, which are anatomically
divorced from events that occur in the superficial
layer (15).
HERMAN ET AL
Recognizing that chondrocytes contain both
acid and neutral proteinases (16,17), attention has
focused in recent years on their potential role in
mediating matrix depletion. Catabolin, a component(s)
in a family of interleukin-1 (IL-1) molecules, is a
product of synovial tissue which induces the synthesis
and release of matrix-degrading proteinases from viable chondrocytes (18,191. IL-1 activity is produced by
a variety of cell types, which include synovial fibroblasts, activated monocyte/macrophages, and monokine-dependent activated T lymphocytes (2CL23). IL-1
has been identified in pathologic synovial fluids
(24,25). Its potential significance in the pathogenesis of
cartilage destruction in rheumatoid arthritis (RA) and
OA remains speculative.
One cannot exclude the production of additional forms of chondrocyte catabolism-inducing activity by pathologic synovium. Based on their capacity to
induce collagenase release from synovial cells and
dermal fibroblasts, macrophage-derived tumor necrosis factor (26) and products of mast cells (27) have such
a potential. The present study was conducted to assess
the effect of select classes of NSAIDs on the synthesis
and activity of catabolism-inducing factors produced
by RA and OA syrlovial tissue.
MATERIALS AND METHODS
Synovial tissue culture. Synovium samples were obtained during total hip or knee replacement from 16 patients
who fulfilled American Rheumatism Association diagnostic
criteria for classic RA (28). Samples were also taken from 15
patients diagnosed as having primary OA, based on the
absence of predisposing secondary factors and the presence
of radiologic evidence of disease. A 72-hour NSAID washout period was instituted prior to surgery. This time period
was selected arbitrarily, since data assessing synovial tissue
levels of various classes of NSAIDs after their discontinuation are unavailable. The administration of steroids,
hydroxychloroquine, chrysotherapy, penicillamine, o r
cytotoxic drugs was not interrupted in the RA group. Tissue
was immediately processed under aseptic conditions. Representative portions were submitted for routine histologic
analysis. The synovium was dissected free of fat and adherent fibrous tissue and was minced into 2-3-mm segments.
Contaminating blood and synovial fluid were removed by
repetitive washing over a 4-hour period in Hanks' balanced
salt solution, pH 7.4, augmented with penicillin (100
units/ml) and streptomycin (100 pg/rnl).
Tissue was maintained in Eagle's minimum essential
medium (EMEM; 10 ml medium per gram of tissue) supplemented with 2 mM L-glutamine, 25 mM HEPES buffer,
antibiotics, and 15% a2-macroglobulin-depleted, decomplemented newborn calf serum. Six-day cultures were established at 37"C, in an atmosphere of 5% C 0 2 and 95% air, in
the presence and absence of piroxicam, 5 pg/ml (lot no.
9603-262-2m; Pfizer, Groton, CT), indomethacin, 1.5 &ml,
or sodium salicylate, 160 &ml. Like medium was exchanged at 72 hours, and that derived from the 2 harvest
periods was pooled. In further experiments to assess the
effect of NSAIDs on the activity of catabolism-inducing
factors, drugs were added to synovium conditioned medium
(SCM) or control medium at the time of harvest. The
concentrations of NSAIDs used corresponded to therapeutic
plasma levels achieved in humans after oral drug administration. Each drug was solubilized in 0.93M sodium carbonate buffer, pH 9.0, and was subsequently diluted with fresh
medium to achieve the desired concentration.
Assay for catabolism-inducingactivity. SCM and control media (media not exposed to synovial tissue, but maintained in the presence or absence of NSAIDs, or media to
which "SAIDs were added at culture termination) were
diluted 1:l with fresh medium and exposed to bovine
weight-bearing articular cartilage discs (4 x 1 mm) which
were prelabeled with 35S04 as described previously (22).
Aliquots of cartilage were repetitively frozen and thawed to
destroy chondrocytes, and were heated to 60°C for 1 hour to
inactivate indigenous enzymes. This nonviable substrate
was used to measure proteoglycan-degrading enzyme activity in SCM, whereas viable cartilage provided an index of
catabolism-inducing activity as well. Lipopolysaccharide (25
pg/ml) was used in assays as a positive control because of its
ability to induce chondrocyte proteinase release (29). Cultures were maintained at 37"C, in 5% COz and 95% air for 6
days. Analogous fresh medium was replaced at 72 hours.
The labeled sulfate present in combined culture supernatants
and explant washes and that remaining in acid-hydrolyzed
explants (6N HCl, lOO"C, 1 hour) was determined by scintillation counting. Catabolism-inducing activity was expressed as the percentage of 3sS04 released from viable
cartilage minus that released from nonviable tissue. Release
from nonviable tissue was negligible (mean 5 SEM 0.37 4
0.28%). Assays were performed in replicates of 5. The
significance of differences between 35S04release occurring
in SCM derived from NSAID-treated and untreated cultures
and that in corresponding control media was determined by
2-tailed t-test.
Effect of piroxicam on synovial tissue protein synthesis. Six-day OA and RA synovial tissue cultures of comparable wet weight were established in leucine-free EMEM
shpplemented as described above, in the presence and
absence of 5 pg/ml of piroxicam. Cultures were continuously
pulsed with 2 pCi/ml of L[4,5-3H]-leucine (New England
Nuclear, Boston, MA). Medium and label were exchanged at
3 days. Culture supernatants were dialyzed free of
unincorporated label, and an aliquot was counted. Tissue
was resuspended in phosphate buffered saline, pH 7.4, and
homogenized (Polytron; Brinkmann Instruments, Westbury,
NY). An aliquot was removed for protein determination
according to the method of Lowry et a1 (30). The clarified
homogenate was precipitated with trichloroacetic acid
(TCA) in a final concentration of 5%. The precipitated
protein was solubilized in 1N NaOH, and an aliquot was
counted. Newly synthesized protein was expressed as the
counts per minute derived from culture supernatants and the
tissue homogenate per gram of tissue protein content.
259
CARTILAGE DESTRUCTION BY NSAID
catabolism-inducing factor was analyzed in 12 cultures. As shown in Figure 2A, piroxicam appeared to
suppress factor synthesis in 11 of the 12 assays. The
postculture addition of drug to SCM blocked activity
in only 3 of these experiments.
In contrast, in 6 assays, neither indomethacin
nor sodium salicylate affected factor synthesis (Figures 2B and C). When these drugs were added to
SCM, catabolism-inducing activity was blocked in 1
experiment in each group.
Rheumatoid arthritis synovium. Sixteen specimens that produced catabolism-inducing activity were
studied. Piroxicam appeared to suppress factor synthesis in 11 of 12 experiments, whereas postculture
addition of drug had a blocking effect in 8 of 11 studies
(Figure 3A).
Indomethacin suppressed factor synthesis and
blocked activity in 2 of 7 experiments (Figure 3B),
whereas comparable effects were observed in only 1 of
6 assays using sodium salicylate (Figure 3C).
Effect of piroxicam on synovial tissue protein
synthesis. Because piroxicam, in contrast to the other
NSAIDs studied, appeared to consistently suppress
catabolism-inducing factor synthesis, experiments
RESULTS
Effect of NSAIDs on cartilage catabolism.
NSAIDs alone had no effect on cartilage proteoglycan
catabolism as reflected by release of labeled sulfate
from viable cartilage during the 6-day culture period.
Representative experiments are shown in Figure 1.
Results using NSAID-containing media cultured for
6-day periods (medium exchanged at 3 days) or media
to which NSAIDs were freshly added at the time
of catabolic assay were comparable. The percentage
of spontaneously released 35S04 differed between
batches of cartilage used. The mean percentage k
SEM of label release, contrasting, respectively, naive
cultures to NSAID-exposed cartilage, was as follows:
2.98 versus 30.3 ? 2.89; indopiroxicam, 31.01
methacin, 31.63 4.1 versus 30.88 f. 4.1; Na salicylate, 31.7 ? 3.9 versus 30.9 f. 3.95.
Osteoarthritic synovium. Fifteen OA specimens
producing catabolism-inducing activity were studied.
The limited quantity of synovial tissue available per
specimen restricted the number of experiments that
would permit simultaneous drug analysis using each of
the 3 classes of NSAIDs investigated. The in vitro
effect of piroxicam on the synthesis and activity of
*
*
60
lndornethactn
(1.5 UQlml)
50
m
-Wm
40
$ 40
30
d
m
2
$
a+
30
0
’
#
20
20
10
5
10
I
2
3
4
5
6
7
9
8
10
I
2
7
8
3
4
5
6
7
8
9
1
0
60
50
W
f 40
-m0
Kc 30
0’
$
20
ip
10
5
1
2
3
4
5
6
9
10
Figure 1. The effect of nonsteroidal antiinflammatory drugs (NSAIDs) on the release of ’%04
from viable
bovine articular cartilage explants in 6-day cultures. Results represent the mean SEM of 5 replicate assays
*
per experiment. Comparable results were obtained whether NSAIDs were maintained in culture for 6 days
(comparable with synovial tissue cultures) prior to their addition to 3sS04-prelabeled cartilage or added
freshly to explants. Prior to assay, control and NSAID-containingmedia were diluted 1 : 1 with fresh medium.
260
HERMAN ET AL
A P
m
35
r
30
-
E
<
30
i
0. Indomethacm
25
.
10-
Y)
SCM Alone
SCM Generated
m Presence
ol Drug
t
-
15
-
10
-
POS-CUI1YrB
Drug
Addition
to SCM
of
5 -
#
01
I
SCM Alone
SCY Qmerated
k R..me
of Drug
POSt-CURUe
AddRlon 01
Drug to SCY
Figure 2. The effect of nonsteroidal antiinflammatory drugs (NSAIDs) on the synthesis (synoviumconditioned medium [SCM] generated in presence of drug) and activity (postculture addition of drug
to SCM) of the catabolism-inducingfactor produced in 6-day explant cultures of osteoarthritic synovial
tissue. Each joined series of data points represents the mean percentage of '304release (assays
performed in replicates of 5 ) greater than corresponding control values (media with or without NSAID)
derived from independent tissue specimens. Closed circles represent results that are significantly
different from the effect of SCM alone.
were performed to exclude a drug toxic effect; total
protein synthesis by synovial tissue cultured in the
presence and absence of the drug was assessed. As
shown in Table 1, piroxicam had no suppressive effect
on the incorporation of 'H-leucine into TCAprecipitable counts in either 6-day OA or RA synovial
tissue cultures, which suggests a selectivity of effect
on catabolism-inducing factor production. The percentage of 3H-leucine-incorporated counts released
into culture supernatants was comparable between
drug-exposed (34.75 2 3.05%) and untreated (34.8 t
4.3%) OA and RA specimens (Table 1). In one experiment using RA synovial tissue, a small, but statistically significant, increase in 3H-leucine incorporation
was observed.
DISCUSSION
Explant cultures of idiopathic OA and RA
synovial tissue have been shown to produce catabolism-inducing activity, as measured by release of
35S04-labeledproteoglycans from viable articular cartilage substrates. The mean percentage of SCMassociated 35S04-proteoglycanrelease that exceeded
control medium values in 6-day cartilage cultures
varied from 19.1 ? 1.8% (range 10.5-34.9) in OA
specimens to 21.7 k 2.3% (range 6.6-36.9) in RA
tissue cultures. The range variance observed in the 16
RA patients may reflect, in part, the medication still
being administered to them at the time of surgery (5-10
mg prednisone, 7 patients; hydroxychloroquine I pa-
CARTILAGE DESTRUCTION BY NSAID
261
35
P
30.
25
-
-
;
20 -
j
15
-
10
-
20-
0
IS-
f
10-
U
I
0 --
-.,
SCM Alone
SCM Generated
m Presence
ot Drug
Post-Culture
Addmon 01
Drua 10 SCM
SCM Alone
35
SCM Generaled
I" presence
of Drug
Past-Culture
Addition 01
Drug to SCM
r
SCM Alone
SCM Qsneraled
Post-Culture
Addition 01
in Prewnce
of Drug
Drug l o SCM
Figure 3. The effect of nonsteroidal antiinflammatory drugs (NSAIDs) on the synthesis (synoviumconditioned medium [SCM] generated in the presence of drug) and activity (postculture addition of
drug to SCM) of the catabolism-inducing factor produced in 6-day explant cultures of rheumatoid
synovial tissue. Each joined series of data points represents the mean percentage of 35S04release
(assays performed in replicates of 5 ) greater than corresponding control values (media with or without
NSAID) derived from independent tissue specimens. Closed circles represent results that are
significantly different from the effect of SCM alone.
Table 1. In vitro effect of piroxicam on synovial membrane protein synthesis*
Synovial tissue
~
Osteoarthritis
Experiment 1
Experiment 2
Rheumatoid arthritis
Experiment 1
Experiment 2
~
~
% difference
Pi'
Control
Piroxicam
118.2 ? 1.7
83.7 2 3.5
118.5 -+ 2.8
90.4 2 1.6
0.25 2 2.4
8.01 f 1.9
0.925
0.107
341.4 f 10.7
199.3 2 1.7
316.5 f 6.6
214.5 2 5.2
-7.3 t 1.9
7.1 f 2.6
0.07
0.021
* Values are the mean f SEM cpm ( X 10-3/gm)of 3H-leucine-incorporatedcounts derived from the sum
of nondialyzable cpm released into culture supernatants and the trichloroacetic acid-precipitable counts
present in tissue homogenates. Three aliquots of synovial tissue of comparable wet weight were used
for each phase per experiment.
t Difference from control medium values as determined by 2-tailed t-test.
262
tient; penicilllamine 1 patient, methotrexate 1 patient;
azathioprine 1 patient). These results are to be contrasted with a mean percentage release of 18.6 2 2.9%
(range 13.3-19.2) produced by SCM derived from 3
normal synovial tissue specimens cultured under identical conditions (data not shown).
We have previously shown that the labeled
proteoglycans released in the presence of such SCM
are heterogeneous in size and that monomers have a
decreased capacity to aggregate with hyaluronic acid
(31). Thus, significant enzymatic cleavage has occurred.
Piroxicam used in vitro in a concentration comparable with a therapeutic serum level obtainable in
humans has been shown to significantly suppress the
synthesis of catabolism-inducing factor(s) by OA and,
possibly, by RA synovial tissue. Because piroxicam
significantly blocked factor activity in the majority of
RA specimens, it remains uncertain whether inhibitory
responses observed with conditioned media generated
in the presence of piroxicam reflected a carryover
effect of drug on factor activity. A suppression of
synthesis in the absence of blocking of activity occurred in only 2 assays.
Experiments assessing the effect of piroxicam
on synovial tissue incorporation of 3H-leucine into
newly synthesized protein indicate that this effect was
not due to drug toxicity. Piroxicam did not inhibit total
protein synthesis, which suggests a selectivity of the
drug-induced suppressive response. Piroxicam alone
had no significant effect on the release of proteoglycans from normal cartilage, further indicating a drugmediated effect at a synovial tissue level.
Indomethacin and sodium salicylate, in the single therapeutic concentrations used, did not affect the
synthesis of catabolism-inducing factor by OA synovium. Occasional suppression occurred in RA tissue
cultures.
N o consistent blocking effect on OA catabolic
factor activity occurred using any of the NSAIDs
studied. However, in contrast with indomethacin and
sodium salicylate, piroxicam more consistently
blocked activity produced by RA synovium. This
suggests that there are potential differences between
the catabolism-inducing factor generated from OA
tissue and that generated from RA tissue, and that
piroxicam may be an effective blocker for that produced by RA synovial tissue. Monocyte IL-1 heterogeneity is recognized (32-34), but there has been little
effort to contrast the quality of that produced by
synovial tissue from patients with different diseases.
HERMAN ET AL
Wood et a1 ( 3 9 , using enzymatically disaggregated OA
and RA synovium, have recently reported a difference
in the isoelectric focusing profiles of IL-1 obtained
from these 2 sources.
Unfractionated SCM was used in the experiments reported here. No attempt was made to identify
the responsible catabolism-inducing cytokine(s). We
have, however, been able to show that comparably
derived OA and RA SCM does contain IL-1 activity,
as assessed using a C3HIHeJ mouse thymocyte proliferation assay. In addition, chromatographic analyses
performed on OA SCM have shown that catabolisminducing activity primarily elutes in a molecular weight
range of 12,000-20,000, consistent with IL-1 (36).
The pathophysiologic significance of IL-1 produced by OA and RA synovial tissue currently remains speculative. IL-1 may have potential deleterious
effects that contribute to cartilage and bone destruction beyond its property of inducing the synthesis and
release of matrix-degrading proteinases from chondrocytes. It is able to stimulate chondrocyte release of
prostaglandins and plasminogen activator (23,37), suppress chondrocyte glycosaminoglycan synthesis (38),
induce the release of prostaglandins, plasminogen activator, and matrix-degrading proteinases from dendritic synoviocytes (39), induce bone resorption
(40,41), and suppress bone collagen and osteocalcin
synthesis (42). Therefore, pharmacologic intervention
to modulate the synthesis and/or activity of IL-1 would
appear to be highly desirable.
Few studies have addressed the effect of
NSAIDs on the synthesis or activity of catabolisminducing factors produced by monocyte/macrophages
or synovial tissue. Ridge et a1 (43) reported that
indomethacin had no effect on such factor synthesis
by lipopolysaccharide-stimulated rabbit peritoneal
macrophages. In contrast, Phadke et a1 (44) observed
that the addition of aspirin and indomethacin (but not
phenylbutazone or fenoprofen) to such macrophageconditioned media could suppress chondrocyte
collagenase and other neutral proteinase release without. affecting cell viability. Why there is a discrepancy
between the latter study and our observation that
indomethacin was essentially ineffectual in blocking
catabolism-inducing activity present in SCM is not
clear. It could perhaps reflect a difference in the
quality of factor(s) produced by the different cell/tissue
sources studied.
Rainsford has recently reported the effect of
various NSAIDs on purified porcine leukocyte IL-1
activity (45). Therapeutic concentrations of diflunisal,
CARTILAGE DESTRUCTION BY NSAID
indomethacin, mefenamic acid, naproxen, piroxicam,
salicylate, and sulindac were without effect. Pharmacologic concentrations of benoxaprofen, diclofenac,
fenclofenac, indomethacin, and sulindac sulfide did
effectively block activity. However, a compromise of
chondrocyte viability was not excluded as a basis for
reduction in IL- 1-induced proteoglycan release from
cartilage. That piroxicam (0.1-500 pM) did not block
the activity of leukocyte-derived IL- 1 is consistent
with our results using OA SCM, but differs from that
employing conditioned media derived from RA tissues. Although the cartilage substrate (bovine nasal)
and duration of cartilage culture (4 days) differed from
our experimental design, results again suggest a possible difference in catabolism-inducing factors produced by the 2 sources studied.
In experiments employing RA synovium cocultured with bovine nasal cartilage (46) and SCM derived
from a rabbit model of antigen-induced arthritis (47),
indomethacin appeared to have no effect on either the
synthesis or activity of catabolism-inducing factors.
Jones et a1 (48), using normal bovine synovium, found
that indomethacin had a variable effect on SCMmediated cartilage breakdown. Sheppeard et a1 (49),
using conditioned medium derived from normal porcine synovium, noted that aspirin failed to affect
catabolism-inducing activity.
The majority of the above studies have failed to
demonstrate the effectiveness of therapeutic concentrations of NSAIDs of diverse classes in modulating
either the synthesis or activity of catabolism-inducing
activity generated from leukocyte or synovial tissue
sources. The current experimental data, however,
indicate that a member of at least 1 class of these
compounds, piroxicam, has the potential capacity to
suppress such factor synthesis by OA and RA
synovium, as well as block the activity of that produced by RA tissue. Sheppeard and Couchman (50)
have recently reported findings that substantiate the
uniqueness of this drug. They observed that the addition of a therapeutic concentration of piroxicam to
normal porcine synovial tissue significantly suppressed SCM-induced cartilage breakdown.
It must be emphasized that the importance of
IL- 1 or other possible catabolism-inducing factors in
the pathogenesis of cartilage destruction in human
disease is presently unknown. Furthermore, caution
must be exercised in extrapolating the significance of
such in vitro organ culture data to in vivo situations.
Nevertheless, the data do suggest that at least 1 class
of NSAID has the potential of modulating a catabo-
263
lism-inducing pathway which perhaps contributes to
cartilage destruction.
ACKNOWLEDGMENTS
We thank Drs. Clark N. Hopson, Edward Miller,
Richard Freiberg, Robert Heidt, Sr., and Edward Zenni, Jr.
for assistance in securing synovial tissue preparations and
Susan Hansen for assistance in manuscript preparation.
REFERENCES
1. Wong S, Gardock JF, Pruss TP: Pharmacologic evaluation of tolectin (tolmetin MCN-2559) and MCN-2891,
two anti-inflammatory agents. J Pharmacol Exp Ther
185:127-138, 1973
2. Ackerrnan NR, Rooks WH 11, Shott L, Genant H,
Maloney P, West E: Effects of naproxen on connective
tissue changes in the adjuvant arthritic rat. Arthritis
Rheum 22:1365-1374, 1979
3. Otterness IG, Larson DL, Lombardino JG: An analysis
of piroxicam in rodent models of arthritis. Agents Actions 12:308-312, 1982
4. Blackham A, Radziwonick H: The effect of drugs in
established rabbit monoarticular arthritis. Agents Actions 7:473-480, 1977
5 . Sloboda AE, Birnbaum JE, Oronsky AL, Kerwar SS:
Studies on type I1 collagen-induced polyarthritis in rats:
effect of antiinflammatory and antirheumatic agents.
Arthritis Rheum 24:616424, 1981
6. Kerwar SS, Sloboda AE: Studies of the effect of
fenbufen and sulindac on type I1 collagen-induced arthritis in rats. Agents Actions 12:216-220, 1982
7. Kerwar SS, Oronsky AL: Type I1 collagen induced
arthritis in rats: clinical and immunological studies. Int J
Tissue React 7:409429, 1985
8. Roach JE, Tomblin W, Eyring EJ: Comparison of the
effect of steroid, aspirin and sodium salicylate on articular cartilage. Clin Orthop 106:35&356, 1975
9. Simmons DP, Chrisman OD: Salicylate inhibition of
cartilage degeneration. Arthritis Rheum 8:96&969, 1965
10. Herman JH, Hess EV: Nonsteroidal anti-inflammatory
drugs and modulation of cartilaginous changes in
osteoarthritis and rheumatoid arthritis. Am J Med
(suppl) 77: 16-25, 1984
11. Brackertz D, Hagmann J, Kueppers F: Proteinase inhibitors in rheumatoid arthritis. Ann Rheum Dis 34:225230, 1975
12. Hadler NM, Johnson AM, Spitznagel JK, Quinet RJ:
Protease inhibitors in inflammatory synovial effusions.
Ann Rheum Dis 4055-59, 1981
13. Cooke TD, Hurd ER, Jasin HE, Bienenstock J, Ziff M:
Identification of immunoglobulins and complement in
rheumatoid articular collagenous tissue. Arthritis
Rheum 18541-551, 1975
14. Jasin HE: Autoantibody specificities of immune com-
HERMAN ET AL
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
plexes sequestered in articular cartilage of patients with
rheumatoid arthritis and osteoarthritis. Arthritis Rheum
281241-248, 1985
Mitchell NS, Shepard N: Changes in proteoglycan and
collagen in cartilage in rheumatoid arthritis. J Bone Joint
Surg 60A:349-354, 1978
Sapolsky AI, Keiser H , Howell DS: Metalloproteases of
human articular cartilage that digest cartilage proteoglycans at neutral and acid pH. J Clin Invest 58: 1030-1041,
1976
Sapolsky AI, Malemud CJ, Norby DP, Moskowitz RW,
Matsuta K, Howell DS: Neutral proteinases from articular chondrocytes in culture. 2. Metal-dependent latent
neutral proteoglycanase and inhibitory activity. Biochim
Biophys Acta 658:138-147, 1981
Saklatvala J, Dingle JT: Identification of catabolin, a
protein from synovium which induces degradation of
cartilage in organ culture. Biochem Biophys Res Commun 96: 1225-123 1, 1980
Saklatvala J, Sarsfield SJ, Pilsworth LMC: Characterization of proteins from human synovium and mononuclear leukocytes that induce resorption of cartilage
proteoglycan in vitro. Biochem J 209:337-344, 1983
Pilsworth LMC, Saklatvala J: The cartilage-resorbing
protein catabolin is made by synovial fibroblasts and its
production is increased by phorbol myristate acetate.
Biochem J 216:481-489, 1983
Jasin HE, Dingle JT: Human mononuclear cell factors
mediate cartilage matrix degradation through chondrocyte activation. J Clin Invest 68571-581, 1981
Herman JH, Greenblatt D, Khosla RC, Appel AM:
Cytokine modulation of chondrocyte proteinase release.
Arthritis Rheum 27:79-91, 1984
McGuire-Goldring MB, Meats JE, Wood DD, Ihrie EJ,
Ebsworth NM, Russell RGG: In vitro activation of
human chondrocytes and synoviocytes by a human
interleukin- 1-like factor. Arthritis Rheum 27:654-662,
1984
Wood DD, Ihrie EJ, Dinarello CA, Cohen PL: Isolation
of an interleukin-1-like factor from human joint effusions. Arthritis Rheum 26:975-983, 1983
Nouri ME, Panayi GS, Goodman SM: Cytokines and the
chronic inflammation of rheumatic disease. 1. The presence of interleukin-1 in synovial fluids. Clin Exp Immunol55:295-302, 1984
Dayer J-M, Beutler B, Cerami A: Cachectinhmor necrosis factor stimulates collagenase and prostaglandin
E2 production by human synovial cells and dermal
fibroblasts. J Exp Med 1622163-2168, 1985
Yoffe JR, Taylor DJ, Wooley DE: Mast cell products
stimulate collagenase and prostaglandin E production by
cultures of adherent rheumatoid synovial cells. Biochem
Biophys Res Commun 122:270-276, 1984
Ropes MW, Bennett GA, Cobb S, Jacox R, Jessar RA:
1958 revision of diagnostic criteria for rheumatoid arthritis. Bull Rheum Dis 9:175-176, 1958
29. Jasin HE: Bacterial lipopolysaccharides induce in vitro
degradation of cartilage matrix through chondrocyte
activation. J Clin Invest 72:20142019, 1983
30. Lowry OH, Rosebrough NJ, Farr AL, Randall RJ:
Protein measurement with the Fohn phenol reagent. J
Biol Chem 193265-275, 1951
31. Appel AM, Trenkamp CA, Bartholomew LE, Hopson
CN, Herman JH: Studies on cartilage matrix depletion
induced in control and pathologic synovial conditioned
media (abstract). Trans Orthop Res SOC11259, 1986
32. Saklatvala J, Sarsfield SJ, Townsend Y: Pig interleukin
1: purification of two immunologically distinct leukocyte
proteins that cause cartilage resorption, lymphocyte
activation, and fever. J Exp Med 162: 1208-1222,
1985
33. Ihrie EJ, Wood DD: Biochemical heterogeneity of human interleukin-1. Lymphokine Res 4: 169-181, 1985
34. Durum SK, Schmidt JA, Oppenheim JJ: Interleukin 1:
an immunological perspective. Annu Rev Immunol
3:263-287, 1985
35. Wood DD, Ihrie EJ, Hamerman D: Release of interleukin-1 from human synovial tissue in vitro. Arthritis
Rheum 28:853-862, 1985
36. Herman JH, Appel AM, Hopson CN: In vitro modulation of cartilage anabolic function by osteoarthritic
synovium (abstract). Arthritis Rheum (suppl) 29:S14,
1986
37. Evequoz V , Schnyder J, Trechsel U, Baggiolini M,
Fleisch H: Influence of macrophage products on the
release of plasminogen activator, collagenase, 0glucuronidase and prostaglandin E2 by articular chondrocytes. Biochem J 219:667-677, 1984
38. Dingle JT: The effect of synovial catabolin on cartilage
synthetic activity. Connect Tissue Res 12:277-286, 1984
39. Mizel SB, Dayer J-M, Krane SM, Mergenhagen SE:
Stimulation of rheumatoid synovial cell collagenase and
prostaglandin production by partially purified lymphocyte activating factor (interleukin 1). Proc Natl Acad Sci
USA 7824742477, 1981
40. Gowen M, Wood DD, Ihrie EJ, McGuire MKB, Russell
RGG: An interleukin 1 like factor stimulates bone resorption in vitro. Nature 306:378-380, 1983
41. Heath JK, Saklatvala J , Merkel MC, Atkinson SJ,
Reynolds JJ: Pig interleukin 1 (catabolin) is a potent
stimulator of bone resorption in vitro. Calcif Tissue Int
37~95-97, 1985
42. Beresford JN, Gallagher JA, Gowen M, Couch M, Poser
J, Wood DD, Russell RGG: The effects of monocyteconditioned medium and interleukin 1 on the synthesis
of collagenous and non-collagenous proteins by mouse
bone and human bone cells in vitro. Biochim Biophys
Acta 801:58-65, 1984
43. Ridge SC, Oronsky AL, Kenvar SS: Induction of the
CARTILAGE DESTRUCTION BY NSAID
265
synthesis of latent collagenase and latent neutral
protease in chondrocytes by a factor synthesized by
activated macrophages. Arthritis Rheum 23:448-454,
1980
Phadke K, Nanda S, Lee K: Release of proteases from
cartilage cells as a result of activation by a macrophage
factor-effects of some antiinflammatory drugs. Biochem Pharmacol 28:3671-3673, 1979
Rainsford KD: Preliminary investigations on the pharmacological control of catabolin-induced cartilage destruction in vitro. Agents Actions 1655-57, 1985
Sledge CB, Steinberg JJ, Tsukamoto S: Communications between synoviocytes and chondrocytes: effect of
hydrocortisone and other agents. Semin Arthritis
Rheum (suppl) 11:136-137, 1981
Kenvar SS, Ridge SC, Landes MJ, Nolan JC, Oronsky
AL: Induction of collagenase synthesis in chondrocytes
by a factor synthesized by inflamed synovial tissue.
Agents Actions 1 4 5 4 5 7 , 1984
48. Jones IL, Klamfeldt A, McGuire MB: Enhanced breakdown of bovine articular cartilage proteoglycans by
conditioned synovial medium. Scand J Rheumatol
11:41L46, 1982
49. Sheppeard H, Pilsworth LMC, Hazelman B, Dingle JT:
Effects of antirheumatoid drugs on the production and
action of porcine catabolin. Ann Rheum Dis 41:463-468,
1982
50. Sheppeard H , Couchman K: Effect of piroxicam,
colchicine and diphosphonate on the production and
action of catabolin from porcine synovial cells. Proceedings of the XVI Congress of Rheumatology, Sydney,
Australia, 1985
44.
45.
46.
47.