Abstract
The effects of slow-release calcium clodronate on rat adjuvant arthritis were investigated using two different dosing schedules. In prophylactic treatment, calcium clodronate was given on the same day as the adjuvant injection, and in therapeutic treatment, calcium clodronate administration was delayed until the animals had active disease, to day 14 postadjuvant. Calcium clodronate was given as single i.m. injections into the thigh muscles. Arthritis index, histopathology of hindpaw, quantitative histomorphometry, bone mineral density and serum osteocalcin, alkaline phosphatase and calcium were studied. Calcium clodronate given therapeutically decreased the severity of paw swelling slightly more than prophylactic treatment, a result seen as lower scores of arthritis index. Histopathological evaluation of hindpaws showed that calcium clodronate protected against inflammation-induced bone loss and reactive bone formation in the hindpaw, but not against inflammatory changes involving articular cartilage. Quantitative histomorphometric analysis of the distal femur indicated that trabecular bone area was decreased by 86% in arthritic rats compared with normal untreated controls. Both the prophylactic and the therapeutic treatment with calcium clodronate prevented this osteopenia (P < .001). Bone mineral density measured by computed tomography was also significantly reduced in distal femoral metaphysis in adjuvant arthritic rats, but restoration to virtually normal values occurred with calcium clodronate (P < .001). In both dosing schedules, we observed a suppression of arthritis, which was associated with a decrease in paw swelling and an inhibition of the severe osteopenia in the distal femoral metaphysis. The long duration of action after a single injection of calcium clodronate indicates that the insoluble salt remains at the injection site and is released slowly into the bloodstream.
The beneficial effect of disodium clodronate administered s.c. in the treatment of arthritic rats has been previously demonstrated in several studies (Flora, 1979; Nugentet al., 1993; Österman et al., 1994a;Österman et al., 1995). Disodium clodronate suppressed the intensity of joint swelling and prevented articular and bone lesions thought to be secondary to inflammation (Östermanet al., 1994a).
Compared with solutions of soluble salts of bisphosphonic acids such as disodium clodronate, suspensions of crystalline insoluble calcium salts, though they are near neutral in pH, reduce tissue damage, local pain and irritation at the injection site. Therefore, the crystalline calcium salt of bisphosphonic acid is especially suitable for i.m. and s.c. administration. In addition, the insoluble salt remains at the injection site and is only slowly released into the bloodstream (Ostovic and Brenner, 1995). This slow systemic release results in the desired concentration of bisphosphonate in the blood for longer periods of time and may thus provide benefits in the treatment and prevention of diseases involving excessive bone resorption.
The adjuvant model represents a systemic inflammatory disease, with bone and cartilage changes similar to those observed in rheumatoid arthritis. The common pathological features of adjuvant arthritis in rat and rheumatoid arthritis in humans are joint swelling associated with cellular and pannus invasion of the joint space, release of lysosomal constituents into the joint space and bone resorption (Pearson and Wood, 1959).
Quantitative histomorphometry has been used to assess more precisely the model of adjuvant arthritis in rat and the effects of drugs on various histopathological features seen in this experimental model (Langman et al., 1990; del Pozo et al., 1990; Jeeet al., 1993; Bonnet et al., 1993). Recently, pQCT has also been used to measure changes in bone mineral density in rats (Turner et al., 1994; Sato et al., 1995).
This study was carried out to investigate the effects of calcium clodronate, administered prophylactically or therapeutically as a single i.m. injection, on bone and articular damage occurring during the progression of adjuvant arthritis.
Materials and Methods
Animals.
Thirty-five male Lewis rats (ZFV Hannover, Germany), mean weight 235 g (S.E.M. 2), were used. At the start of the study, the animals were 6 to 7 weeks old. They were fed a standard pellet diet (SDS RM3 E SQC, Witham, UK) and had free access to tap water during the study. The temperature of the animal room was maintained at 21 ± 2°C, and the relative humidity at 50 ± 15%. The illumination of the room was artificially controlled, with cycles of 12 h of light and 12 h of darkness.
Induction of adjuvant arthritis.
The animals were given an intradermal injection of heat-killed mycobacteria (Mycobacterium tuberculosis, Ministry of Agriculture and Fishery, Central Veterinary Laboratories, Weybridge, Surrey, UK) suspended in light mineral oil (Pristane, Sigma St. Louis, MO). The bacteria were first ground in a mortar to fine powder. The powder was weighed and suspended in pristane (2 mg/ml). The suspension was dispersed evenly just before intra-dermal injection of 0.1 ml at the base of the tail.
Ethical guidelines for experimental investigations in animals were followed, and every effort was made to minimize pain (Zimmerman, 1983).
Administration of the test compounds and measurement of clodronate in serum.
The day of adjuvant injection was defined as day 0. In the prophylactic group (n = 10), the animals received micronized (Neo Micro AG, Switzerland, grain size < 4 μm) calcium clodronate (Leiras Oy, Finland) suspended in saline (250 mg/ml) on day 0 as single i.m. injection (0.5 ml/kg) into the thigh muscles (125 mg/kg/thigh, total dose 250 mg/kg). In the therapeutic group (n = 10), an equal dose of calcium clodronate was given as single i.m. injection on day 14 postadjuvant. The arthritic control group (n = 10) received single injections of saline. Before therapeutic administration, rats with similar scores were assigned to the arthritic control group and the therapeutic group. The animals were sacrificed on day 28 after immunization. Five untreated animals served as normal nonarthritic controls.
Calcium clodronate (250 mg/kg) was given as single i.m. injection (0.5 ml/kg) to a separate group of animals [Sprague-Dawley rats weighing 245 (SD 10) g]. Blood samples were collected at days 1, 2, 4, 7, 10 and 14. Three animals were used for each time-point, and the concentration of clodronate in serum was analyzed by capillary gas chromatography with mass-selective detection (Österman et al., 1994b).
Evaluation of clinical signs of arthritis.
The rats were weighed, and the arthritis index was scored blindly by the same person for each rat once a week. The hindpaws, forepaws, ears, nose and tail were graded separately, according to the severity of redness and swelling. The severity of arthritis of each hindpaw was scored as 0 = no swelling, 1 = minimal swelling, 2 = medium swelling and 3 = severe swelling. Each forepaw in which swelling was present in at least one joint was scored 0.5. Connective tissue swelling in the nose was scored 1, presence of knots in the tail 1 and each ear in which redness appeared and knots had formed 0.5. The maximal score for each animal was 10.
Biochemical assessment.
At the time of sacrifice, blood samples were collected by cardiac puncture under anesthesia with CO2 (CO2:O2 = 1:1). Serum was separated by centrifugation and stored at −20°C or −80°C (for osteocalcin determination). Serum alkaline phosphatase activity and calcium were determined colorimetrically (Boehringer Mannheim). Serum osteocalcin level was measured by specific radioimmunoassay based on rat osteocalcin (Biomedical Technologies, Inc., Mannheim, Germany).
Sample collection for bone histology and densitometry.
The right femur and hindpaw were removed at necropsy and processed non-decalcified for histology (Schenk et al., 1984). The left femur was removed, cleaned from soft tissue and stored at 4°C in 70% ethanol until used for determination of bone mineral density. The distal end of the right femur was cut sagittally with a precision bone saw (Exakt cutting grinding system, Exakt Apparatebau, Norderstedt, Germany). The hindpaw was cut above the tarsus, including the distal tibia along with the talus, calcaneus, distal row of tarsal bones, metatarsals and surrounding soft tissue. According to our preliminary studies, the arthritic scores of the right and left hindpaws of each animal were virtually the same.
Preparation of specimens for histopathological evaluation and bone histomorphometry.
Longitudinal 5-μm-thick sections of the lateral part of distal femoral metaphysis and hindpaw specimen were cut in a standardized sagittal plane with a Polycut S heavy-duty microtome (Reichert-Jung, Leica Instruments GmbH, Germany). The specimens were stained using the von Kossa method with hematoxylin/McNeal’s tetrachrome counterstain (Schenk et al., 1984), and hindpaw specimens were also stained using the Masson-Goldner trichrome method (Schenk et al., 1984). Histopathological evaluation of the stained hindpaw sections was done with light microscopy.
Bone histomorphometry.
Static histomorphometric measurement of the distal femoral specimens was carried out using a MCID digital image analyzer with M1 morphometry software (Imaging Research Inc., St. Catharines, Ontario, Canada). Measurement was carried out on the primary and secondary spongiosa of the distal femoral metaphysis at 4× objective magnification. The first microscopic field (field 1) of femoral metaphysis was located within 0.3 to 1.7 mm and the second field (field 2) within 1.7 to 3.1 mm of the growth cartilage-metaphyseal junction. A mean total metaphyseal tissue area of 5.52 mm2 was measured for each femur. Total tissue area, total mineralized trabecular bone area and total bone perimeter were measured fully automatically, and the percentage of total mineralized trabecular bone area, trabecular thickness, number and separation were calculated according to Parfitt et al. (1983).
Bone densitometry with pQCT.
Bone mineral density of the distal femur and mid-diaphysis of the femur were measured by pQCT (XCT 960A, Stratec, Germany) using special software for densitometric and geometric studies in small animals. To control the stability of measurement, the instrument was calibrated every day against a phantom of known mineral content. Before scanning, the length of the excised left femur was measured with a micrometer, and the femur was then placed anterior side up in the measuring tube. Three 1-mm-thick sections of the distal femur at 0.8-mm intervals were scanned using a voxel size of 0.148 × 0.148 × 1.0 mm. The distance of the middle section from the distal end of the femur was 5 mm. This area corresponds approximately to the sections of the femur on which histomorphometric measurements were performed. The total bone mineral density (milligrams per cubic centimeter) of the sections was measured. The mineral density of cortical bone was analyzed at the mid-diaphysis of the femur. The coefficient of variation for the measurement of bone density by pQCT was 1% in our laboratory.
Statistical analysis.
All statistical analyses were accomplished using SAS system (SAS Institute, Inc., Cary, NC, ver. 6.10, 1994). Bone mineral density and biochemical variables were analyzed with one-way analysis of variance. Comparisons between groups were performed using contrasts. The statistical analysis of the arthritis index at 2, 3 and 4 weeks after immunization was done using repeated-measures analysis of variance with between factor “group” and within factor “week.” If there was a significant group × week interaction, comparison between the groups was done using interaction contrasts. Static histomorphometric variables were analyzed using repeated-measures analysis of variance with between factor “group” and within factor “field.” If there was a significant group × field interaction, one-way analysis of variance was performed separately at both fields, and comparison between the groups was done using linear contrasts. All comparisons were carried out with Bonferroni adjustment. Numerical values are given as mean and SEM. A P value lower than .05 was considered statistically significant.
Results
Concentration of clodronate in serum.
The concentration of clodronate in serum after single i.m. injections of calcium clodronate was followed for 2 weeks. Clearly measurable concentrations in serum were found as late as 14 days after administration, which indicated that clodronate is released slowly from the injection site. Serum levels were 1558 (S.E.M. 228) μg/l, 626 (40) μg/l, 420 (22) μg/l, 355 (60) μg/l, 216 (32) μg/l and 167 (29) μg/l at days 1, 2, 4, 7, 10 and 14, respectively.
Clinical variables.
Immunization of Lewis rats with mycobacteria in oil gave rise to rapid development of arthritis in every rat. Clinical onset of symptoms was seen 14 days after immunization. Arthritic rats showed a progressive increase in arthritis index, which peaked 21 days postadjuvant and began to decrease thereafter. A decrease in weight gain was seen during the disease process. At the end of the study, the mean weight in the normal untreated group was 318 (S.E.M. 4) g and in the arthritic control group was 226 (4) g (P < .001). In therapeutic treatment, however, calcium clodronate, decreased significantly the loss of body weight when compared with the arthritic controls [248 (7) g vs.226 (4) g, P < .05].
Fourteen days after immunization, the arthritis index (scale 1–10) was 6.2 (S.E.M. 0.4) in the arthritic control group, 4.6 (0.6) in the prophylactic treatment group and 5.2 (0.7) in the therapeutic treatment group. In all groups, the arthritis index increased from week 2 to week 3 and decreased from week 3 to week 4. However, no significant difference was found between the groups in the rise or the fall of the arthritis index. At the end of the study, the lowest arthritis index, 5.5 (0.4), was found in the therapeutic treatment group and the highest 8.3 (0.4), in the arthritic control group.
Serum biochemical variables.
After the administration of calcium clodronate to arthritic rats, some changes were observed in biochemical variables (table 1). At the time of sacrifice (day 28), serum alkaline phosphatase activity was significantly lower in the arthritic control group and the prophylactic group, and serum calcium level was significantly lower in the arthritic control group, as compared with normal untreated controls. Serum osteocalcin level was reduced (P < .001) by calcium clodronate under both dosing schedules in comparison with normal untreated and arthritic controls.
Bone densitometry.
A 49% lower total bone mineral density in distal femoral metaphysis was observed in the arthritic control group on day 28 than in normal untreated animals. Both prophylactic and therapeutic calcium clodronate treatments prevented this decrease; bone mineral density in both groups was significantly higher than that in arthritic controls but was not significantly different from that in normal controls. In addition, bone density was significantly higher in the prophylactically treated group than in the therapeutically treated group (figs. 1 and 2).
A slight but significant reduction in cortical bone mineral density of the femur measured at the level of mid-diaphysis was observed in arthritic rats. However, this reduction was not prevented by calcium clodronate treatment (fig. 2).
Bone histomorphometry.
In field 1, the percentage of mineralized trabecular bone area was 27.2% in normal untreated rats and 3.7% in arthritic control rats (table 2). This corresponds to a 86% decrease in trabecular bone area in arthritic control rats compared with normal untreated rats. The trabecular osteopenia was due to a significant decrease in trabecular thickness and number, leading to an increase in trabecular separation (table 2). Both the prophylactic and the therapeutic treatments with calcium clodronate completely prevented the development of trabecular osteopenia in the distal femoral metaphysis, maintaining the percentages of total mineralized bone area at the level observed in normal untreated rats. Both treatment schedules with calcium clodronate also prevented the decrease in trabecular number (P < .001) and the increase in trabecular separation (P < .001) but had no significant effect on trabecular thickness in field 1. Furthermore, prophylactic treatment increased the trabecular number above the level in untreated control rats (P < .01).
In field 2, trabecular bone area was very low in normal untreated rats compared with field 1, and no significant trabecular osteopenia was induced by adjuvant arthritis in this region (table 2). Prophylactically and therapeutically administered calcium clodronate had different effects (P < .001). Although both treatment schedules significantly increased trabecular bone area as compared with the arthritic control rats, prophylactic treatment also increased trabecular bone area above the level in untreated rats (P < .001). In addition, in prophylactically treated rats, trabecular thickness and number were significantly higher, and trabecular separation significantly lower, than in arthritic or normal untreated rats (table 2). The histological findings of the distal femur are shown in figure 3.
Histopathological evaluation of the hindpaw.
In the arthritic control group, the dermis and other soft tissues were heavily invaded by inflammatory cells—mainly lymphocytes and other mononuclear cells, but also some granulocytes. Synovial tissue showed a marked inflammatory reaction and hyperplasia, and in many areas articular cartilage was removed and destroyed by pannus tissue. Several granuloma-like formations were seen both in surrounding soft tissue and, especially, associated with synovial tissue. There were also several intraosseous granuloma-like formations, which destroyed mainly trabecular bone in the tibiotarsal region. Marked increase was seen in bone resorption as well as in bone formation. The reactive bone formation in the periosteal regions largely destroyed the normal outlines of bone, and in some cases even ectopic bone formation could be seen (fig. 4B).
In the arthritic group treated prophylactically with calcium clodronate, bone resorption was clearly diminished, and trabecular bone loss was not nearly so pronounced as in the arthritic control group. There was no clear evidence of other effects of treatment compared with the arthritic control group (fig. 4C).
In the arthritic group treated therapeutically with calcium clodronate, synovial tissue showed marked hyperplasia, and an inflammatory reaction was seen in most joint spaces, but no intraosseous granulomas were observed and the trabecular bone structure was intact. Also, the reactive bone formation along the periosteal surface was more moderate in this group than in the arthritic control and prophylactic groups (fig. 4D).
Discussion
Our previous study of continuous administration of disodium clodronate to adjuvant arthritic rats at a dosage of 25 mg/kg/day over a 2-week period (administered five times a week, cumulative dose 250 mg/kg) indicated a suppression of the intensity of joint swelling and prevention of secondary inflammatory articular and bone lesions in the tibiotarsal region (Österman et al., 1994a). In the present study, we obtained similar results for the inhibition of paw inflammation and bone changes after single i.m. injections of calcium clodronate (250 mg/kg) to arthritic rats.
A disadvantage of bisphosphonic acids in pharmaceutical applications is that they cause tissue damage and localized pain and irritation after i.m. or s.c. injection. Another disadvantage is that the level of bisphosphonic acid in the serum reaches a peak within an hour and disappears from the serum rapidly thereafter (Conrad and Lee, 1981;Yakatan et al., 1982; Laurén et al., 1991;Österman et al., 1994b). Administration of insoluble calcium salts of bisphosphonates offers at least three advantages over soluble bisphosphonates. They reduce tissue damage and pain on injection, provide slow systemic release of the bisphosphonic acid (Ostovic and Brenner, 1995) and allow administration at long intervals. Slow release of clodronate was also found in this study after administration of calcium clodronate.
Immunization of Lewis rats with mycobacteria rapidly induced severe arthritis. This vigorous inflammation leads also to systemic effects such as reduction in weight gain. It is interesting that therapeutic treatment with slow-release clodronate decreased the weight loss, which suggests that it could somehow affect the systemic consequences of inflammation. Arthritic rats showed a progressive increase in arthritis index, which peaked on day 21 and begun to decrease thereafter. Treatment with calcium clodronate, particularly therapeutic treatment, reduced the inflammation as reflected in the arthritis index. Histopathological evaluation of the hindpaws revealed that in the arthritic control group, there were marked inflammatory reactions with granulomas and pannus formation, which resulted in marked destruction of articular cartilage and bone in the tibiotarsal region. Calcium clodronate treatment protected against inflammation-induced secondary bone loss and granuloma formation as well as reactive bone formation in the hindpaw, but not against the inflammatory changes involving the synovium and articular cartilage. The effect was more beneficial after therapeutic treatment, a result that may be due to a more suitable scheduling of administration during the disease process.
In studies using an adjuvant arthritis model, bone changes in the hindpaw have been evaluated (Flora, 1979; Francis et al., 1989). Recent studies conducted in arthritic rats have shown generalized osteopenia in addition to local demineralization (del Pozoet al., 1990; Bonnet et al., 1993; del Pozo and Zapf, 1994). Although the rat adjuvant arthritis model includes severe synovitis in the tibiotarsal joint region in the hindpaw, no marked synovitis has been found in the knee joint region. Despite the lack of synovitis and soft-tissue inflammation in the knee joint, very severe trabecular bone loss was observed in the distal femur of arthritic control rats. Although we had no base-line controls (Kalu, 1991), it is obvious that true bone loss occurred. As shown by histomorphometric studies, this osteopenia was effectively prevented by single i.m. injections of calcium clodronate given either prophylactically or therapeutically. However, the effects of the two treatment schedules on structural indices of metaphyseal trabeculae differed. An excessive increase in mineralized trabecular bone area in the primary and secondary spongiosa was found after prophylactic treatment. An accumulation of trabecular bone was seen also in the region more distant from the growth plate because of longitudinal bone growth in the 2-month-old rats and the antiresorptive effect of the drug. Thus the prophylactic treatment indicated a long duration of action (4 weeks) after a single dose of calcium clodronate.
del Pozo and Zapf (1994) have shown that bone density is a sensitive parameter in experimental arthritis. Our results obtained by pQCT showed the ability of calcium clodronate to maintain bone mineral density in distal femoral metaphysis in adjuvant arthritic rats. In agreement with the histomorphometric studies, an excessive increase in bone mineral density was noticeable after prophylactic treatment. However, a slight but significant decrease in cortical bone mineral density in femoral midshaft, observed in arthritic rats, was not prevented by calcium clodronate treatment.
The best-known consequences of rheumatoid arthritis are inflammation of the synovium and destruction of cartilage. Moreover, rheumatoid arthritis is characterized by both periarticular and generalized osteoporosis (Sambrook and Reeve, 1988; Joffe and Epstein, 1991). Therefore, antiresorptive drugs such as clodronate might be advantageous in the treatment of rheumatoid arthritis. Besides its effective inhibition of bone resorption, clodronate had a beneficial effect on inflammation as shown by decreased arthritis index. Bisphosphonates seem to have a direct effect on cells involved in inflammatory reaction. Earlier studies have shown that liposome encapsulated clodronate is phagocytosed by macrophages and induces a depletion of macrophages in the liver and the spleen (Mönkkönen et al., 1991; Camilleri et al., 1995). This is probably due to the tendency of clodronate to induce apoptosis in osteoclasts and macrophages after phagocytosis (Selander et al., 1996 in press). Further studies are now needed to investigate whether bisphosphonates, given as some kind of an insoluble complex such as calcium clodronate or in liposomes locally to the arthritic joint region, can be used to eliminate macrophages in the synovium and thus reduce inflammation and cartilage damage. Kinneet al. (1995) have recently shown a long-term amelioration of rat adjuvant arthritis after systemic administration of clodronate-containing liposomes. This amelioration was parallelled by an elimination of activated macrophages in immunocompetent areas of the spleen and draining lymph nodes.
In conclusion, our results indicate that calcium clodronate, given as single i.m. injections to adjuvant arthritic rats, reduces the intensity of joint swelling and secondary bone lesions in the tibiotarsal region of the hindpaw. In addition, the severe osteopenia in the distal femoral metaphysis of arthritic rats was effectively inhibited by calcium clodronate treatment. The long duration of action after single injection of calcium clodronate indicates that the insoluble salt remains at the injection site and is slowly released into the bloodstream.
Acknowledgments
We would like to thank the staff of Preclinical Research at Leiras Oy for expert technical assistance.
Footnotes
-
Send reprint requests to: Thua Österman, Biomedical Research Center, Leiras Oy, P.O. Box 415, FIN-20101 Turku, Finland.
- Abbreviations:
- pQCT
- peripheral quantitative computed tomography
- MCID
- microcomputer imaging device
- Received June 7, 1996.
- Accepted October 15, 1996.
- The American Society for Pharmacology and Experimental Therapeutics