Prevention BYEmilio
of Adjuvant Arthritis
by Cyclosporine
in Rats
del Pozo, Peter Elford, Romain Perrelet, Michael Graeber, Jean Paul Casez,
Dominique
Modrowski,
Trevor Payne, and Andrew
The effect of cyclosporine A during the development phase of adjuvant arthritis was studied in 40 female rats. Five groups of eight animals each received oral cyclosporine, 2.5, 5, 10, 20, or 30 mg/kg daily for 30 days. Also, eight normal and eight diseased rats served as placebo controls. At the time of inoculation of the adjuvant suspension on day 0, measurement of disease parameters (paw swelling and vertebral density) was started concomitantly with beginning of therapy. On completion of the study, the animals were killed, and after measurement of total skeletal and segmental (hind legs and caudal spine plus two caudal vertebrae) calcium, the two assessed vertebrae and both femoral condyles were removed for histomorphometric evaluation (vertebrae) and for estimation of glycosaminoglycan (GAG) content of cartilage. Blood for osteocalcin determinations also was taken at term from control and untreated arthritic rats and from animals that had received 10 mg/kg cyclosporine. Treatment with 2.5 mg/ kg was ineffective, but doses between 5 and 20 mg/kg prevented the development of articular and osseous lesions. The 20 mg/kg dose showed no
T
HE SPECTRUM of action of cyclosporine A has been investigated widely in experimental animals and in humans, but information on its effect on bone remodeling is limited. This compound has been found to counteract the in vitro release of Ca’+ from bone explants induced by different agents, and evidence recently has been provided that this effect is mediated via suppression of immunological phenomena.‘-3 In vivo studies have shown cyclosporine to have protective effects on osteopenia accompanying established experimental arthritis.4 Studies in nonarthritic rats have been equivocal: whereas some studies report a positive action of cyclosporine on bone,5-7 others show a loss of trabecular tissue under similar experimental conditions.x.’ This report presents additional findings concerning the action of cyclosporine on the skeleton during the developmental phase of experimental articular inflammation.
Roland MacKenzie
better effect than 10 mg/kg. This was shown by the absence of inflammation and the presence of normal condylar GAG and total mineral content in the areas screened. Untreated animals showed marked reductions in all of these parameters. The 30 mg/kg dose was effective in blocking the GAG loss, but significant reductions in bone density and trabecular volume were seen. There was a close correlation between GAG and bone density values, suggesting a common causal relationship. Circulating osteocalcin was significantly elevated in the untreated animals with adjuvant arthritis. The administration of 10 mg/kg cyclosporine reduced this parameter considerably but not to basal levels. In conclusion, cyclosporine effectively prevented the development of experimental arthritis and associated bone mineral loss in oral doses of 5 to 10 mg/kg; 20 mg/kg showed no better effect than IO mg/kg, and the 30 mg/kg dose produced toxic skeletal demineralization. Copyright
(~8 1992
by W.B. Saunders
Company
INDEX WORDS: Cyclosporine; developing arthritis; bone; adjuvant arthritis. received oral cyclosporine
or vehicle, 2.5. 5, 10. 20, or 30 mg/kg daily for 30 days. Eight normal rats served as controls. Adjuvant was injected on day 0 concomitantly with beginning of therapy. Upon completion of the study, the animals were killed and caudal vertebrae 2 and 3 and femoral
MATERIALS AND METHODS
Forty-eight female rats were randomly assigned to six groups of eight animals each. Each group Seminars in Arthritis and Rheumatism,
Vol 2 1, No 6, Suppl 3 (June), 1992: pp 23-29
23
24
condyles were removed for histomorphometry and glycosaminoglycan (GAG) determinations, respectively. Jugular blood was collected at the same time. The following variables were investigated. Articular Swelling Hind-paw diameter was assessed at 5-day intervals using a specially designed micrometric caliper. Only values recorded on days 15 and 30 are presented.
DEL PO20
ET AL
of arthritic rats receiving 10 mg/kg cyclosporine were compared with those of normal controls and untreated arthritic animals. Determinations were conducted using a double photon source of gamma radiation and a specially designed collimator for rat bone (Hologic Inc, Waltham, MA). In addition to total skeletal measurements, selective scanning was performed on hind legs and caudal vertebrae. Osteocalcin Determinations
Bone Densitometry Densitometry measurements of the two first caudal vertebrae were performed by a previously described method.” In brief, the procedure involves densitometric scanning of standardized radiographs of caudal vertebrae using a high-resolution computerized imaging device (Imaging Research, Ontario, Canada). With this device values are automatically processed into relative optical density (ROD) units. The mean trabecular density was selected as a suitable variable because previous studies have shown a close correlation (y = 1.Olx + 0.036; r = 0.99) between cortical and trabecular values. Results are expressed as the tangent of the calculated regression line (ROD v time) of values recorded between days 1 and 30 (n = 8 rats/group).
On completion of treatment-only in the control and arthritic groups and in the group receiving 10 mg/kg cyclosporine-plasma osteocalcin was measured by a specific rat radioimmunoassay. I2 Statistical evaluation was accomplished by analysis of variance or Student’s t tests. Results are expressed as mean + SE.
Bone Morphometry
Vertebral Density
Caudal vertebrae 2 and 3 were dissected from control rats, animals with untreated arthritis, and groups receiving 10 and 30 mg/kg cyclosporine daily. The vertebrae were embedded in methylmethacrylate, and 6-pm sections were cut with a Jung-K microtome (Reichert-Jung, Heidelberg, Germany). Subsequently, von Kossa’s stains of the sections were evaluated for total trabecular volume, and resorption (osteoclastic lacunae) surfaces were expressed as the percentage of trabecular area using computer-assisted image integration.
As expected, rats with arthritis had lower vertebral density values than controls at the end of the treatment period (Fig 2A). Cyclosporine doses of 5 and 20 mg/kg restored vertebral densities to normal, while 10 mg/kg produced density levels significantly (P < .Ol) higher than those seen in control animals and in animals treated with 5 and 20 mg/kg doses. The 2.5 mg/kg dose was ineffective, as previously reported.4 In contrast, a significant (P < .Ol) decrease in density was recorded with the 30 mg/kg dose.
Cartilage Measurements After the cartilage had been digested with papain, a direct spectrometric assay was used to estimate GAG levels in whole femoral condyles removed postmortem.” Total Skeletal Calcium At the end of treatment, total skeletal calcium values (as grams of hydroxyapatite) in the group
RESULTS
Paw Swelling Cyclosporine treatment prevented paw swelling in a dose-related manner. The 2.5 mg dose had no effect, but doses of 5 mg/kg and above showed a clear protective effect (P < .O1 for 5 mg and P < .OOl for 10 to 30 mg). Values recorded at 15 and 30 days are shown in Fig 1.
Histomorphometry Vertebral trabecular volume was significantly (P < .Ol) reduced in the animals with arthritis receiving no treatment and in rats receiving 30 mg/kg (P < .02) (Fig 3A). Compared with control rats, this finding was associated with a substantial increase (P < .OOl and P < .Ol for rats with arthritis and rats receiving 30 mg/kg, respectively) in osteoclastic resorption (Fig 3B) and in perios-
PREVENTION OF ADJUVANT
ARTHRITIS
Paw diameter (mm) Day 15
8
6
10
Day 30
l-
8
6
25
BY CYCLOSPORINE
numerous interspersed osteoclasts. In general, these multinucleated cells showed positive acid phosphatase staining indicative of an activated state. Abnormal vascular proliferation was conspicuous within the bone marrow undergoing fibrotic degeneration. Treatment with 10 mg/kg cyclosporine prevented bone destruction and maintained parameters in the range of control animals (Fig 4E). The 30 mg/kg dose resulted in significantly (P < .05) reduced trabecular volume and enhanced (P < .05) bone resorption in accordance with the distinct density decrease reported in the previous section (Fig 3).
In general, GAG content of femoral condyles paralleled density profiles (Fig 2B). Thus, tangent values of radiometric estimations closely correlated with condylar content of GAG (Y = .75; P < .OOl) (Fig 5). This correlation indicates the presence of a link between chondral and osseous processes. After induction of adjuvant arthritis, 10 mg/kg of cyclosporine daily provided the
A i
i
c
AA 2.5
5
SIM Fig 1:
10
20
30 mg/kg +
Progression of paw swelling on days 15
and 30 of treatment
with cyclosporine [SIM, cy-
closporine (Sandimmune”)].
There is significant (P
< .Ol) inhibition at 5 mg/kg and complete block-
GAG
ade (P < .OOl) at 210 mg/kg. C, control; AA, un-
bl)
treated adjuvant arthritis.
teocytic lacunae formation (Fig 4A and B). Some trabecular surfaces were covered largely by unmineralized osteoid embedding “lost” osteoblasts (Fig 4C). In some areas. intense mineral resorption had led to replacement of bone structures by a fibrotic mass (Fig 4D). The microscopic characteristics of trabecular remnants were consistent with nonfunctional, so-called woven bone. Compared with those of control animals, the bone marrow spaces were hypercellular, often occupied almost exclusively by fibroblasts with
600 500 400 300 200 100 0i
Fig 2:
B
_
i Adjd2.5 SIM
20
30 mglkg l
(A) Density profiles expressed as the tan-
gent of the regression line composed of ROD units in dependency of time (days 1 to 30). (B) Profile of glycosaminoglycan values. SIM, cyclosporine (Sandimmunem).
26
DEL POZO
was reduced in untreated animals with arthritis. This finding is consistent with previous vertebral trabecular bone data4 and indicates that bone loss is systemic. Treatment with cyclosporine restored mineral content to the normal range. Selective measurements in the area affected by the inflammatory process (hind legs and caudal vertebrae) showed enhancement of mineral apposition under cyclosporine therapy and provided further evidence of this substance’s positive effect on bone remodeling.
Trabecular volume (%) 30
A
T
25
ET AL
T
20 T
15 10 5
Osteocalcin Determinations
0i
Compared with controls, plasma osteocalcin was significantly elevated in untreated animals with arthritis (Table 2). Treatment with 10 mg/kg cyclosporine considerably reduced circulating osteocalcin without reaching basal levels. This probably is a result of the low-level continuation of bone repair.
Resorption surface (%) 30.
-I-
25 20.
DISCUSSION
15 10
5 0~
II C
AA
21: Fig 3:
30 mglkg SIM
(A) Vertebral trabecular volumes as per-
centile of total volume. Values are significantly reduced in the untreated
animals with arthritis (P
< .Ol) and at the 30 mg/kg dose (P < -02). (B) Osteoclastic
resorption lacunae as percentile
of
total trabecular surface. Animals with arthritis and rats receiving
30 mg/kg
had larger resorption
areas (P < .OOl and P < .Ol, respectively) than controls. SIM,
cyclosporine
(Sandimmune@).
C,
control; AA, untreated adjuvant arthritis.
maximal protective effect on bone and cartilage. The 20 mg/kg dose resulted in a tendency toward lower GAG values. Total Skeletal Calcium Total-body calcium values obtained in the animals receiving 10 mg/kg cyclosporine are presented in Table 1. Total skeletal mineral content
Initiation of treatment with cyclosporine at the time of inoculation of the adjuvant suspension prevented the development of articular inflammation in a dose-related manner. The 2.5 mg/ kg dose previously had been found ineffective in treating established experimental arthritis4 This dose also did not block development of the disease when administered prophylactically, whereas doses of 5 to 20 mg/kg were fully effective. The results illustrated in this report are in agreement with recently published data. Thus, daily cyclosporine doses of 5 to 15 mg/kg administered to rats with established adjuvant arthritis were sufficient to promote regression of osseous and artitular signs of inflammation within 10 days, as shown by radiometric, biochemical, and histological studies.4 However, although articular swelling was completely prevented in the present study, cyclosporine treatment reduced paw diameter considerably in animals with established disease without achieving values recorded in normal controls.4 Several aspects of the present work deserve further comment. Enhanced mineral density recorded at the 10 mg/kg dose may be attributable to improved growth velocity in response to therapy. Indeed, recent studies conducted in this laboratory have shown a close correlation between bone mineral content and skeletal growth (data
PREVENTION
Fig 4:
OF ADJUVANT
ARTHRITIS
Histomorphometry
becular structures.
BY CYCLOSPORINE
of healthy and diseased bone (original magnification
x160).
(A) Normal tra-
Few erosions are visualized; there is no noticeable osteocytic activity, and the ap-
pearance of bone marrow is normal. (B) Severe osteoclastic bone resorption and accompanying activity of intratrabecular and become hypercellular.
osteocytes
hyper-
producing confluent lacunae. Bone marrow has lost fat droplets
(C) Failure of osteoid mineralization
despite enhanced osteoblastic activity.
There is severe periosteocytic resorption. (D) Fibrous degeneration of bone marrow and almost complete demineralization
of the remaining trabeculae.
(E) Protective effect of 10 mg/kg cyclosporine. There are
few repair areas. Osteoclasts are not visualized, and there is no apparent osteocytic activity. Bone marrow appears normal.
_
28
DEL POZO ET AL
2o011 Fig 5:
@F)
Correlation between
bone density values
(ROD) and condylar GAG content.
not presented). Although the 30 mg/kg dose fully prevented paw swelling, the optical density of vertebral bone was clearly diminished concomitant with a reduction in trabecular volume. These findings may be attributable to a direct effect of cyclosporine on mineral mobilization from the skeleton because electron microscopic examination of the parathyroid glands has failed to show enhanced hormone synthesis as the possible cause. Bone histomorphometry confirmed the favorable effect of cyclosporine on the inflammatory process. Thus, the loss of trabecular volume and enhanced osteoclastic activity characteristic of arthritic osteopenia were normalized by cyclosporine therapy. Partial normalization of circulating osteocalcin levels (as a parameter of new
bone formation) was also seen. This finding may indicate that bone repair continues even when histomorphometric parameters are apparently restored to normal. In addition, local signs of inflammation (eg, bone marrow infiltration and vascular proliferation) were absent after cyclosporine therapy. Vertebral densitometry and total skeletal calcium measurements were in agreement with histological data, reflecting generalized skeletal involvement in the arthritic process and restoration of normal bone remodeling by cyclosporine. The positive effect of cyclosporine on bone in experimental arthritis is at variance with the results of some reported studies in normal animals. In normal rats, administration of cyclosporine has been found to enhance skeletal remodeling and induce bone loss as shown by tibia1 histomorphometry and sequential osteocalcin determinations.8 When cyclosporine was discontinued, partial recovery was noted.’ The compound also was found to enhance bone resorption in the oophorectomized rat, whereas simultaneous administration of prednisone seemed to minimize the adverse bone effects of either agent given alone.‘3”4 More recently, it was reported that simultaneous administration of cyclosporine and 1,2Sdihydroxyvitamin D3 can restore bone loss induced by cyclosporine alone.” These findings contrast with results of other in vitro and in vivo experiments. Thus, cyclosporine inhibited bone resorption in cultured mouse calvaria.16 In another study, cyclosporine had no effect on Ca2+ release from bone explants but was found to counteract the release of calcium induced by parathyroid hormone, vitamin D, and
Table 1: Total Skeletal CaZ+ Hind Legs
Total Ca*+ Rats
(9)
Controls
9.5 * 0.3
l
+ Caudal Spine
Caudal Vertebrae
(9)
2 and 3 (mg)
3.8 + 0.2
139&12
8.6 * 0.1t
3.5 -t O.l$
126+
9.4 * 0.3
4.1 !I 0.25
142+20
Arthritis Untreated Treated,
cyclosporine,
10 mg/kg
NOTE. Data represent mean f SE. * As hydroxyapatite. t P < .03, untreated arthritis v control. $ P < .05, untreated arthritis Y control. § P < .04, treated arthritis v untreated arthritis.
16
PREVENTION
OF ADJUVANT
ARTHRITIS
Table 2: Osteocalcin,
BY CYCLOSPORINE
Day 30
::I_.~~
NOTE Data presented * P i
as nanograms
per m~llihter (mean k SE).
001 (v control)
t P i ,002 (v control).
interleukin 1.’ More recently, suppression of bone resorption by cyclosporine in untreated rats was also reported.5 In fact, cyclosporine enhanced vertebral bone apposition, as estimated by dynamic histomorphometry after double tetracycline labeling. Recent studies conducted in our laboratory have shown that cyclosporine treatment results in moderate demineralization of tibiae but has no such effect on vertebrae.6 These anomalies may result from the different composition of tibia1 and vertebral marrow as the source of cells that influence bone kinetics. Thus, examination of bone marrow composition in tibiae shows an absence of fat and marked hy-
29
percellularity with equal occurrence of mononuclear and segmented forms in addition to polynuclear cells. In contrast, trabecular interspaces of vertebrae are occupied by numerous fat droplets, which are interspersed with a few, seldom multinuclear. cells. In this context. it has been shown that ovariectomy in dogs. known to modify bone mineral content, induces alterations in fat and cellular components of bone marrow.” Consequently, the response of bone to cyclosporine may be dependent on the particular site in the skeleton. Nevertheless, a decrease in bone resorption and osteoclast formation was described recently in rat long bones exposed to cyclosporine.’ This finding lends further support to a protective action of this compound against bone resorption. However, in experimental arthritis this may depend on the biological substrate presented to the pharmacological agent. Most hndings point to an immunologically mediated mechanism because cytokine-dependent inflammatory changes are readily counteracted by cyclosporine, covering any other possible direct effect on bone.
REFERENCES I. Stewart PJ. Green OC. Stem PH: Cyclosporin A inhibits calccmic hormone-induced bone resorption in vitro. J Bone Miner Res I:285291. 1986 2. Sk.jodt H. Crawford A. Elford PR, et al: Cyclosporin A modulates interleukin- I activity on bone in vitro. Br J Rheumatol 24: 165 169. 1985 3. Connolly KM. Stecher VJ, Danis E, et al: Alteration of interleukin-1 activity and the acute phase response following medication of adjuvant arthritic rats treated with cyclosporin A or methotrexate. Int J Immunopharmacol 10:717-728, 1988 3. del Pozo E, Graeber M, Elford P. et al: Regression of bone and cartilage loss in adjuvant arthritic rats after treatment with cyclosporin A. Arthritis Rheum 33:247-252, 1990 5. Orcel P, Bielakoff J, Modrowski D, et al: Cyclosporin A induced in vivo inhibition of resorption and stimulation of formation in rat bone. J Bone Miner Res 4:387-391, 1989 6. del Pozo E, Elford P, Casez JP, et al: Effect of Sandimmune on bone remodelling in normal rats and under various pathological conditions. in Mizushima Y, Amor B (eds): Autoimmunity, Rheumatoid Arthritis andCyclosporin A. Carnforth. England. Parthenon. 1990, pp 53-65 7. Orcel P. Denne MA, de Vernejoul MC: Cyclosporin A in vitro decreases bone resorption, osteoclast formation, and the fusion of cells of the monocyte-macrophage lineage. Endocrinology 128:1638-1646, 1991 8. Movsowitz C, Epstein S, Fallon M, et al: CyclosporinA in v,ivo produces severe osteopenia in the rat: effect ofdose and duration of administration. Endocrinology 123:257 l2577. 1988
9. Schlosbetg M. Movsowitz C. Epstein S. et al: The effect of cyclosporin A administration and its withdrawal on bone mineral metabolism in the rat. Endocrinology 124:2 179-2 184. I989 IO. del Pozo E. Gubler HU, Perrelet R, et al: Non-invasive quantitative estimation of bone density in rats throughout the life cycle and in arthritic osteopenia: Preliminary results. Hormone Metab Res 20:630-632, 1988 1I. Farndale RW, Sayers CA, Barrett AJ: A direct spectrophotometric microassay for sulfated glycosaminoglycans in cartilage cultures. Connect Tissue Res 9:247-248. I982 12. Modrowski D, del Pozo E. Miravet L: Dynamics of circulating osteocalcin in rats under physiological and experimental conditions. Hormone Metab Res (in press) 13. Movsowitz C, Epstein S. Ismail F, et al: Cyclosporin A in the oophorectomised rat: Unexpected severe bone resorption. J Bone Miner Res 4:393-398. 1989 14. Movsowitz C, Schlosberg M, Epstein S, et al: Combined treatment with cyclosporin A and cortisone acetate minimizes the adverse bone effects of either agent alone. J Orthop Res 8:635-64 I, I990 15. Epstein S, Schlosberg M, Fallon M, et al: I.25 Dihydroxyvitamin D3 modifies cyclosporine-induced bone loss. Calcif Tissue Int 47: 152- 157. 1990 16. Klaushofer K, Hoffman 0, Stewart J. et al: Cyclosporin A inhibits bone resorption in cultured neonatal mouse calvaria. J Pharmacol Exp Ther 243:584-590, I987 17. Martin RB, Chow BD, Lucas PA: Bone marrow fat content in relation to bone remodeling and serum chemistry in intact and ovariectomized dogs. Calcif Tissue Int 46: I89194. 1990
of Adjuvant Arthritis
by Cyclosporine
in Rats
del Pozo, Peter Elford, Romain Perrelet, Michael Graeber, Jean Paul Casez,
Dominique
Modrowski,
Trevor Payne, and Andrew
The effect of cyclosporine A during the development phase of adjuvant arthritis was studied in 40 female rats. Five groups of eight animals each received oral cyclosporine, 2.5, 5, 10, 20, or 30 mg/kg daily for 30 days. Also, eight normal and eight diseased rats served as placebo controls. At the time of inoculation of the adjuvant suspension on day 0, measurement of disease parameters (paw swelling and vertebral density) was started concomitantly with beginning of therapy. On completion of the study, the animals were killed, and after measurement of total skeletal and segmental (hind legs and caudal spine plus two caudal vertebrae) calcium, the two assessed vertebrae and both femoral condyles were removed for histomorphometric evaluation (vertebrae) and for estimation of glycosaminoglycan (GAG) content of cartilage. Blood for osteocalcin determinations also was taken at term from control and untreated arthritic rats and from animals that had received 10 mg/kg cyclosporine. Treatment with 2.5 mg/ kg was ineffective, but doses between 5 and 20 mg/kg prevented the development of articular and osseous lesions. The 20 mg/kg dose showed no
T
HE SPECTRUM of action of cyclosporine A has been investigated widely in experimental animals and in humans, but information on its effect on bone remodeling is limited. This compound has been found to counteract the in vitro release of Ca’+ from bone explants induced by different agents, and evidence recently has been provided that this effect is mediated via suppression of immunological phenomena.‘-3 In vivo studies have shown cyclosporine to have protective effects on osteopenia accompanying established experimental arthritis.4 Studies in nonarthritic rats have been equivocal: whereas some studies report a positive action of cyclosporine on bone,5-7 others show a loss of trabecular tissue under similar experimental conditions.x.’ This report presents additional findings concerning the action of cyclosporine on the skeleton during the developmental phase of experimental articular inflammation.
Roland MacKenzie
better effect than 10 mg/kg. This was shown by the absence of inflammation and the presence of normal condylar GAG and total mineral content in the areas screened. Untreated animals showed marked reductions in all of these parameters. The 30 mg/kg dose was effective in blocking the GAG loss, but significant reductions in bone density and trabecular volume were seen. There was a close correlation between GAG and bone density values, suggesting a common causal relationship. Circulating osteocalcin was significantly elevated in the untreated animals with adjuvant arthritis. The administration of 10 mg/kg cyclosporine reduced this parameter considerably but not to basal levels. In conclusion, cyclosporine effectively prevented the development of experimental arthritis and associated bone mineral loss in oral doses of 5 to 10 mg/kg; 20 mg/kg showed no better effect than IO mg/kg, and the 30 mg/kg dose produced toxic skeletal demineralization. Copyright
(~8 1992
by W.B. Saunders
Company
INDEX WORDS: Cyclosporine; developing arthritis; bone; adjuvant arthritis. received oral cyclosporine
or vehicle, 2.5. 5, 10. 20, or 30 mg/kg daily for 30 days. Eight normal rats served as controls. Adjuvant was injected on day 0 concomitantly with beginning of therapy. Upon completion of the study, the animals were killed and caudal vertebrae 2 and 3 and femoral
MATERIALS AND METHODS
Forty-eight female rats were randomly assigned to six groups of eight animals each. Each group Seminars in Arthritis and Rheumatism,
Vol 2 1, No 6, Suppl 3 (June), 1992: pp 23-29
23
24
condyles were removed for histomorphometry and glycosaminoglycan (GAG) determinations, respectively. Jugular blood was collected at the same time. The following variables were investigated. Articular Swelling Hind-paw diameter was assessed at 5-day intervals using a specially designed micrometric caliper. Only values recorded on days 15 and 30 are presented.
DEL PO20
ET AL
of arthritic rats receiving 10 mg/kg cyclosporine were compared with those of normal controls and untreated arthritic animals. Determinations were conducted using a double photon source of gamma radiation and a specially designed collimator for rat bone (Hologic Inc, Waltham, MA). In addition to total skeletal measurements, selective scanning was performed on hind legs and caudal vertebrae. Osteocalcin Determinations
Bone Densitometry Densitometry measurements of the two first caudal vertebrae were performed by a previously described method.” In brief, the procedure involves densitometric scanning of standardized radiographs of caudal vertebrae using a high-resolution computerized imaging device (Imaging Research, Ontario, Canada). With this device values are automatically processed into relative optical density (ROD) units. The mean trabecular density was selected as a suitable variable because previous studies have shown a close correlation (y = 1.Olx + 0.036; r = 0.99) between cortical and trabecular values. Results are expressed as the tangent of the calculated regression line (ROD v time) of values recorded between days 1 and 30 (n = 8 rats/group).
On completion of treatment-only in the control and arthritic groups and in the group receiving 10 mg/kg cyclosporine-plasma osteocalcin was measured by a specific rat radioimmunoassay. I2 Statistical evaluation was accomplished by analysis of variance or Student’s t tests. Results are expressed as mean + SE.
Bone Morphometry
Vertebral Density
Caudal vertebrae 2 and 3 were dissected from control rats, animals with untreated arthritis, and groups receiving 10 and 30 mg/kg cyclosporine daily. The vertebrae were embedded in methylmethacrylate, and 6-pm sections were cut with a Jung-K microtome (Reichert-Jung, Heidelberg, Germany). Subsequently, von Kossa’s stains of the sections were evaluated for total trabecular volume, and resorption (osteoclastic lacunae) surfaces were expressed as the percentage of trabecular area using computer-assisted image integration.
As expected, rats with arthritis had lower vertebral density values than controls at the end of the treatment period (Fig 2A). Cyclosporine doses of 5 and 20 mg/kg restored vertebral densities to normal, while 10 mg/kg produced density levels significantly (P < .Ol) higher than those seen in control animals and in animals treated with 5 and 20 mg/kg doses. The 2.5 mg/kg dose was ineffective, as previously reported.4 In contrast, a significant (P < .Ol) decrease in density was recorded with the 30 mg/kg dose.
Cartilage Measurements After the cartilage had been digested with papain, a direct spectrometric assay was used to estimate GAG levels in whole femoral condyles removed postmortem.” Total Skeletal Calcium At the end of treatment, total skeletal calcium values (as grams of hydroxyapatite) in the group
RESULTS
Paw Swelling Cyclosporine treatment prevented paw swelling in a dose-related manner. The 2.5 mg dose had no effect, but doses of 5 mg/kg and above showed a clear protective effect (P < .O1 for 5 mg and P < .OOl for 10 to 30 mg). Values recorded at 15 and 30 days are shown in Fig 1.
Histomorphometry Vertebral trabecular volume was significantly (P < .Ol) reduced in the animals with arthritis receiving no treatment and in rats receiving 30 mg/kg (P < .02) (Fig 3A). Compared with control rats, this finding was associated with a substantial increase (P < .OOl and P < .Ol for rats with arthritis and rats receiving 30 mg/kg, respectively) in osteoclastic resorption (Fig 3B) and in perios-
PREVENTION OF ADJUVANT
ARTHRITIS
Paw diameter (mm) Day 15
8
6
10
Day 30
l-
8
6
25
BY CYCLOSPORINE
numerous interspersed osteoclasts. In general, these multinucleated cells showed positive acid phosphatase staining indicative of an activated state. Abnormal vascular proliferation was conspicuous within the bone marrow undergoing fibrotic degeneration. Treatment with 10 mg/kg cyclosporine prevented bone destruction and maintained parameters in the range of control animals (Fig 4E). The 30 mg/kg dose resulted in significantly (P < .05) reduced trabecular volume and enhanced (P < .05) bone resorption in accordance with the distinct density decrease reported in the previous section (Fig 3).
In general, GAG content of femoral condyles paralleled density profiles (Fig 2B). Thus, tangent values of radiometric estimations closely correlated with condylar content of GAG (Y = .75; P < .OOl) (Fig 5). This correlation indicates the presence of a link between chondral and osseous processes. After induction of adjuvant arthritis, 10 mg/kg of cyclosporine daily provided the
A i
i
c
AA 2.5
5
SIM Fig 1:
10
20
30 mg/kg +
Progression of paw swelling on days 15
and 30 of treatment
with cyclosporine [SIM, cy-
closporine (Sandimmune”)].
There is significant (P
< .Ol) inhibition at 5 mg/kg and complete block-
GAG
ade (P < .OOl) at 210 mg/kg. C, control; AA, un-
bl)
treated adjuvant arthritis.
teocytic lacunae formation (Fig 4A and B). Some trabecular surfaces were covered largely by unmineralized osteoid embedding “lost” osteoblasts (Fig 4C). In some areas. intense mineral resorption had led to replacement of bone structures by a fibrotic mass (Fig 4D). The microscopic characteristics of trabecular remnants were consistent with nonfunctional, so-called woven bone. Compared with those of control animals, the bone marrow spaces were hypercellular, often occupied almost exclusively by fibroblasts with
600 500 400 300 200 100 0i
Fig 2:
B
_
i Adjd2.5 SIM
20
30 mglkg l
(A) Density profiles expressed as the tan-
gent of the regression line composed of ROD units in dependency of time (days 1 to 30). (B) Profile of glycosaminoglycan values. SIM, cyclosporine (Sandimmunem).
26
DEL POZO
was reduced in untreated animals with arthritis. This finding is consistent with previous vertebral trabecular bone data4 and indicates that bone loss is systemic. Treatment with cyclosporine restored mineral content to the normal range. Selective measurements in the area affected by the inflammatory process (hind legs and caudal vertebrae) showed enhancement of mineral apposition under cyclosporine therapy and provided further evidence of this substance’s positive effect on bone remodeling.
Trabecular volume (%) 30
A
T
25
ET AL
T
20 T
15 10 5
Osteocalcin Determinations
0i
Compared with controls, plasma osteocalcin was significantly elevated in untreated animals with arthritis (Table 2). Treatment with 10 mg/kg cyclosporine considerably reduced circulating osteocalcin without reaching basal levels. This probably is a result of the low-level continuation of bone repair.
Resorption surface (%) 30.
-I-
25 20.
DISCUSSION
15 10
5 0~
II C
AA
21: Fig 3:
30 mglkg SIM
(A) Vertebral trabecular volumes as per-
centile of total volume. Values are significantly reduced in the untreated
animals with arthritis (P
< .Ol) and at the 30 mg/kg dose (P < -02). (B) Osteoclastic
resorption lacunae as percentile
of
total trabecular surface. Animals with arthritis and rats receiving
30 mg/kg
had larger resorption
areas (P < .OOl and P < .Ol, respectively) than controls. SIM,
cyclosporine
(Sandimmune@).
C,
control; AA, untreated adjuvant arthritis.
maximal protective effect on bone and cartilage. The 20 mg/kg dose resulted in a tendency toward lower GAG values. Total Skeletal Calcium Total-body calcium values obtained in the animals receiving 10 mg/kg cyclosporine are presented in Table 1. Total skeletal mineral content
Initiation of treatment with cyclosporine at the time of inoculation of the adjuvant suspension prevented the development of articular inflammation in a dose-related manner. The 2.5 mg/ kg dose previously had been found ineffective in treating established experimental arthritis4 This dose also did not block development of the disease when administered prophylactically, whereas doses of 5 to 20 mg/kg were fully effective. The results illustrated in this report are in agreement with recently published data. Thus, daily cyclosporine doses of 5 to 15 mg/kg administered to rats with established adjuvant arthritis were sufficient to promote regression of osseous and artitular signs of inflammation within 10 days, as shown by radiometric, biochemical, and histological studies.4 However, although articular swelling was completely prevented in the present study, cyclosporine treatment reduced paw diameter considerably in animals with established disease without achieving values recorded in normal controls.4 Several aspects of the present work deserve further comment. Enhanced mineral density recorded at the 10 mg/kg dose may be attributable to improved growth velocity in response to therapy. Indeed, recent studies conducted in this laboratory have shown a close correlation between bone mineral content and skeletal growth (data
PREVENTION
Fig 4:
OF ADJUVANT
ARTHRITIS
Histomorphometry
becular structures.
BY CYCLOSPORINE
of healthy and diseased bone (original magnification
x160).
(A) Normal tra-
Few erosions are visualized; there is no noticeable osteocytic activity, and the ap-
pearance of bone marrow is normal. (B) Severe osteoclastic bone resorption and accompanying activity of intratrabecular and become hypercellular.
osteocytes
hyper-
producing confluent lacunae. Bone marrow has lost fat droplets
(C) Failure of osteoid mineralization
despite enhanced osteoblastic activity.
There is severe periosteocytic resorption. (D) Fibrous degeneration of bone marrow and almost complete demineralization
of the remaining trabeculae.
(E) Protective effect of 10 mg/kg cyclosporine. There are
few repair areas. Osteoclasts are not visualized, and there is no apparent osteocytic activity. Bone marrow appears normal.
_
28
DEL POZO ET AL
2o011 Fig 5:
@F)
Correlation between
bone density values
(ROD) and condylar GAG content.
not presented). Although the 30 mg/kg dose fully prevented paw swelling, the optical density of vertebral bone was clearly diminished concomitant with a reduction in trabecular volume. These findings may be attributable to a direct effect of cyclosporine on mineral mobilization from the skeleton because electron microscopic examination of the parathyroid glands has failed to show enhanced hormone synthesis as the possible cause. Bone histomorphometry confirmed the favorable effect of cyclosporine on the inflammatory process. Thus, the loss of trabecular volume and enhanced osteoclastic activity characteristic of arthritic osteopenia were normalized by cyclosporine therapy. Partial normalization of circulating osteocalcin levels (as a parameter of new
bone formation) was also seen. This finding may indicate that bone repair continues even when histomorphometric parameters are apparently restored to normal. In addition, local signs of inflammation (eg, bone marrow infiltration and vascular proliferation) were absent after cyclosporine therapy. Vertebral densitometry and total skeletal calcium measurements were in agreement with histological data, reflecting generalized skeletal involvement in the arthritic process and restoration of normal bone remodeling by cyclosporine. The positive effect of cyclosporine on bone in experimental arthritis is at variance with the results of some reported studies in normal animals. In normal rats, administration of cyclosporine has been found to enhance skeletal remodeling and induce bone loss as shown by tibia1 histomorphometry and sequential osteocalcin determinations.8 When cyclosporine was discontinued, partial recovery was noted.’ The compound also was found to enhance bone resorption in the oophorectomized rat, whereas simultaneous administration of prednisone seemed to minimize the adverse bone effects of either agent given alone.‘3”4 More recently, it was reported that simultaneous administration of cyclosporine and 1,2Sdihydroxyvitamin D3 can restore bone loss induced by cyclosporine alone.” These findings contrast with results of other in vitro and in vivo experiments. Thus, cyclosporine inhibited bone resorption in cultured mouse calvaria.16 In another study, cyclosporine had no effect on Ca2+ release from bone explants but was found to counteract the release of calcium induced by parathyroid hormone, vitamin D, and
Table 1: Total Skeletal CaZ+ Hind Legs
Total Ca*+ Rats
(9)
Controls
9.5 * 0.3
l
+ Caudal Spine
Caudal Vertebrae
(9)
2 and 3 (mg)
3.8 + 0.2
139&12
8.6 * 0.1t
3.5 -t O.l$
126+
9.4 * 0.3
4.1 !I 0.25
142+20
Arthritis Untreated Treated,
cyclosporine,
10 mg/kg
NOTE. Data represent mean f SE. * As hydroxyapatite. t P < .03, untreated arthritis v control. $ P < .05, untreated arthritis Y control. § P < .04, treated arthritis v untreated arthritis.
16
PREVENTION
OF ADJUVANT
ARTHRITIS
Table 2: Osteocalcin,
BY CYCLOSPORINE
Day 30
::I_.~~
NOTE Data presented * P i
as nanograms
per m~llihter (mean k SE).
001 (v control)
t P i ,002 (v control).
interleukin 1.’ More recently, suppression of bone resorption by cyclosporine in untreated rats was also reported.5 In fact, cyclosporine enhanced vertebral bone apposition, as estimated by dynamic histomorphometry after double tetracycline labeling. Recent studies conducted in our laboratory have shown that cyclosporine treatment results in moderate demineralization of tibiae but has no such effect on vertebrae.6 These anomalies may result from the different composition of tibia1 and vertebral marrow as the source of cells that influence bone kinetics. Thus, examination of bone marrow composition in tibiae shows an absence of fat and marked hy-
29
percellularity with equal occurrence of mononuclear and segmented forms in addition to polynuclear cells. In contrast, trabecular interspaces of vertebrae are occupied by numerous fat droplets, which are interspersed with a few, seldom multinuclear. cells. In this context. it has been shown that ovariectomy in dogs. known to modify bone mineral content, induces alterations in fat and cellular components of bone marrow.” Consequently, the response of bone to cyclosporine may be dependent on the particular site in the skeleton. Nevertheless, a decrease in bone resorption and osteoclast formation was described recently in rat long bones exposed to cyclosporine.’ This finding lends further support to a protective action of this compound against bone resorption. However, in experimental arthritis this may depend on the biological substrate presented to the pharmacological agent. Most hndings point to an immunologically mediated mechanism because cytokine-dependent inflammatory changes are readily counteracted by cyclosporine, covering any other possible direct effect on bone.
REFERENCES I. Stewart PJ. Green OC. Stem PH: Cyclosporin A inhibits calccmic hormone-induced bone resorption in vitro. J Bone Miner Res I:285291. 1986 2. Sk.jodt H. Crawford A. Elford PR, et al: Cyclosporin A modulates interleukin- I activity on bone in vitro. Br J Rheumatol 24: 165 169. 1985 3. Connolly KM. Stecher VJ, Danis E, et al: Alteration of interleukin-1 activity and the acute phase response following medication of adjuvant arthritic rats treated with cyclosporin A or methotrexate. Int J Immunopharmacol 10:717-728, 1988 3. del Pozo E, Graeber M, Elford P. et al: Regression of bone and cartilage loss in adjuvant arthritic rats after treatment with cyclosporin A. Arthritis Rheum 33:247-252, 1990 5. Orcel P, Bielakoff J, Modrowski D, et al: Cyclosporin A induced in vivo inhibition of resorption and stimulation of formation in rat bone. J Bone Miner Res 4:387-391, 1989 6. del Pozo E, Elford P, Casez JP, et al: Effect of Sandimmune on bone remodelling in normal rats and under various pathological conditions. in Mizushima Y, Amor B (eds): Autoimmunity, Rheumatoid Arthritis andCyclosporin A. Carnforth. England. Parthenon. 1990, pp 53-65 7. Orcel P. Denne MA, de Vernejoul MC: Cyclosporin A in vitro decreases bone resorption, osteoclast formation, and the fusion of cells of the monocyte-macrophage lineage. Endocrinology 128:1638-1646, 1991 8. Movsowitz C, Epstein S, Fallon M, et al: CyclosporinA in v,ivo produces severe osteopenia in the rat: effect ofdose and duration of administration. Endocrinology 123:257 l2577. 1988
9. Schlosbetg M. Movsowitz C. Epstein S. et al: The effect of cyclosporin A administration and its withdrawal on bone mineral metabolism in the rat. Endocrinology 124:2 179-2 184. I989 IO. del Pozo E. Gubler HU, Perrelet R, et al: Non-invasive quantitative estimation of bone density in rats throughout the life cycle and in arthritic osteopenia: Preliminary results. Hormone Metab Res 20:630-632, 1988 1I. Farndale RW, Sayers CA, Barrett AJ: A direct spectrophotometric microassay for sulfated glycosaminoglycans in cartilage cultures. Connect Tissue Res 9:247-248. I982 12. Modrowski D, del Pozo E. Miravet L: Dynamics of circulating osteocalcin in rats under physiological and experimental conditions. Hormone Metab Res (in press) 13. Movsowitz C, Epstein S. Ismail F, et al: Cyclosporin A in the oophorectomised rat: Unexpected severe bone resorption. J Bone Miner Res 4:393-398. 1989 14. Movsowitz C, Schlosberg M, Epstein S, et al: Combined treatment with cyclosporin A and cortisone acetate minimizes the adverse bone effects of either agent alone. J Orthop Res 8:635-64 I, I990 15. Epstein S, Schlosberg M, Fallon M, et al: I.25 Dihydroxyvitamin D3 modifies cyclosporine-induced bone loss. Calcif Tissue Int 47: 152- 157. 1990 16. Klaushofer K, Hoffman 0, Stewart J. et al: Cyclosporin A inhibits bone resorption in cultured neonatal mouse calvaria. J Pharmacol Exp Ther 243:584-590, I987 17. Martin RB, Chow BD, Lucas PA: Bone marrow fat content in relation to bone remodeling and serum chemistry in intact and ovariectomized dogs. Calcif Tissue Int 46: I89194. 1990
