in the prevention and treatment of - induced osteoporosis

W.F. Lems1, J.W.G. Jacobs2, J.W.J. Bijlsma2

1Academic Hospital Vrije Universiteit, ABSTRACT osteoclastogenesis are tightly coupled Slotervaart Hospital, Jan van Breemen Since the use of fluoride has been shown (4). Bone morphogenetic proteins pro- Institute, Amsterdam; 2University Medical to stimulate bone formation and since vide the tonic baseline control of both Center, Utrecht, The Netherlands. decreased bone formation is a key fea- processes, which may be influenced by W.F. Lems, MD, PhD, Rheumatologist; ture in the pathogenesis of corticoster- other inputs (e.g. hormones, disease ac- J.W.G. Jacobs, MD, PhD, Rheumatologist; oid-induced osteoporosis (CIOP), fluo- tivity, mechanical strain). J.W.J. Bijlsma, MD, PhD, Professor of ride is, at least theoretically, attractive In histomorphometric studies, a reduced Rheumatology. for the prevention and treatment of CIOP. thickness of osteoid seams and decreased Please address correspondence and In postmenopausal women positive ef- calcification rate, measured by tetracy- reprint requests to: W.F. Lems, Academic fects of low-dose fluoride on the bone cline labeling, have been shown during Hospital Vrije Universiteit, Department of density (BMD) of the lumbar treatment with Cs (5), which indicates Rheumatology (B417), PO Box 7057, 1007 MB Amsterdam, The Netherlands. spine and on the vertebral fracture rate that bone formation is inhibited during E-mail: [email protected] were found; in contrast, in patients treat- treatment with Cs. Recently, it also has Clin Exp Rheumatol 2000; 18 (Suppl. 21): ed with high dose fluoride, an increase been shown that Cs not only have a sup- S65-S68. in the peripheral fracture rate was found. pressive effect on osteoblastogenesis, but Randomized controlled trials of the ef- also promote the apoptosis of osteoblasts © Copyright CLINICAL AND fects of fluoride in Cs-treated patients are and osteocytes (4). EXPERIMENTAL RHEUMATOLOGY 2000. scarce. Although positive effects of low- Biochemical markers are very suitable dose fluoride on BMD of the lumbar to measure changes in bone metabolism. Key words: Fluoride, osteoporosis, spine have been observed, fluoride does In a study in healthy volunteers, it was . not represent the first choice therapy for observed that Cs, even when used in low the prevention or treatment of CIOP, be- dosages, e.g. 10 mg prednisone/day, may cause no positive effects on BMD of the have an inhibiting effect on bone forma- hips and (so far) no reduction in the ver- tion (6). tebral fracture rate have been shown. Data on the influence of Cs on bone re- sorption are somewhat conflicting. In Introduction histomorphometric studies, the number In order to understand the rationale for of osteoclasts and erosion depths were the use of fluoride in (Cs)- increased, indicating that bone resorption treated patients, the pathogenesis of cor- is stimulated in Cs-treated patients (5). ticosteroid-induced osteoporosis (CIOP) Increased bone resorption can be related and the effects of fluoride on bone (me- not only to the use of Cs, but also to the tabolism) will be briefly discussed. Since underlying disease: e.g. in active RA data on the effects of fluoride in the pre- bone resorption is increased (7). With- vention and treatment of CIOP are out interference of the underlying dis- scarce, we will first go into the results ease, in a study in healthy volunteers, of studies with fluoride in postmenopau- urinary excretion of pyridinolines (mark- sal women, and will then discuss the ers of bone resorption) was unchanged available data on the effect of fluoride or even slightly decreased in Cs-treated in the prevention and treatment of CIOP. patients (6). Thus, these data suggest an uncoupling Effects of on bone in Cs-treated patients between bone for- Although it is well known that the use mation (decreased) and bone resorption of corticosteroids (Cs) is associated with (absolutely or relatively increased); the osteoporosis (1-3), the pathogenesis of most important effect of the use of Cs CIOP has not fully been elucidated. Be- seems to be the inhibition of bone for- fore discussing the mechanism of CIOP, mation. it is important to realize that in healthy adults bone remodelling is a continuous Effects of fluoride on bone process, in which osteoblastogenesis and At the molecular level, it is suggested

S-65 Fluoride treatment for glucocorticoid-induced osteoporosis / W.F. Lems et al. that fluoride inhibits a unique fluoride- occur almost entirely in the weight-bear- it seems attractive to use drugs based on sensitive phosphotyrosine phosphatase ing bones, and which resemble the stress the pathophysiology of the underlying (PTP) in osteoblasts, resulting in an in- fractures caused by skeletal overload in disease and the molecular mode of ac- crease in the tyrosine phosphorylation athletes (13). tion of the prescribed drugs. level key signaling proteins of the MAPK Gastrointestinal symptoms are also im- As described, the pathogenesis of Cs- mitogenic transduction pathway, thus po- portant side effects of fluoride; induced osteoporosis is different from tentiating the bone cell proliferation ini- it is thought that these symptoms are elic- that of postmenopausal osteoporosis. In tiated by growth factors (8). ited by a direct irritating effect of fluo- postmenopausal osteoporosis, bone re- In general, studies in which fluoride was ride on the gastric mucosa. Gastrointes- sorption and, to a lesser degree, bone for- used in one of the treatment arms have tinal side effects occur less frequently mation are generally increased, so that shown a strong positive effect on the bone when enteric-coated sodium fluoride tab- anti-resorptive therapy (e.g., with bis- mineral density (BMD) of the spine, lets or monofluorophosphate are used. phosphonates) seems to be a logical strat- while the effects of fluoride at more cor- In a recent study (14), Ringe et al. com- egy. tical sites (e.g. the hips) are much smaller pared the effects of low dose intermit- In contrast, fluoride could be effective

(9, 10). It is thought that this difference tent monofluorophosphate (11.2 mg fluor- against Cs-induced osteoporosis, since is related to the fact that the spine largely ide/day) with continuous (high dose) the inhibition of bone formation plays a consists of trabecular bone, which is monofluorophosphate (20 mg/day) and major role in the pathogenesis of Cs-in- highly active metabolically, while bone with calcium alone: the incidence of new duced osteoporosis (4-6) and since fluo- turnover is much slower at more corti- vertebral fractures per 100 years of treat- ride is the only available pharmacologic cal sites. ment was 8.6, 17.0 and 31.6, respective- agent that has been consistently shown ly. The same continuous regimen of fluo- to stimulate osteoblast activity and bone Effects of fluoride in patients not ride treatment was used in another study formation (10). treated with corticosteroids. (15), in which new vertebral fractures The positive effects of fluoride on BMD Since data on the effects of fluoride in were observed in 2.4% and 10% of pa- in Cs-treated patients have been assessed the prevention and treatment of CIOP are tients treated with monofluorphosphate in retrospective, open and non-random- limited, it is useful to summarize the data or calcium (only), respectively. These ized trials (19-22). In a retrospective stu- on the use of fluoride in postmenopau- results not only indicate that fluoride has dy in which 13 Cs-treated patients were sal (not Cs-treated) osteoporotic women. a preventive effect on fractures in post- taking fluoride (40-60 mg/day sodium In the 1980s, an increase in BMD of the menopausal osteoporotic women, but fluoride), bone mineral mass measured lumbar spine and a decrease in the ver- also illustrate that low dose fluoride may by neutron activation analysis increased tebral fracture rate were observed in pa- be more effective than high dose fluo- from 0.65 to 0.81 after 4 years (19). In tients treated with 40-60 mg sodium fluo- ride in the prevention of osteoporotic an open study of 19 Cs-treated patients, ride daily (11, 12). In another study (13), fractures. the mean increase in BMD at the lum- an increase in the BMD of the lumbar Thus, there seems to be a narrow thera- bar spine was 18.7% after 14 months of spine was accompanied by an increase peutic window for fluoride, since the treatment with 20-30 mg slow-release in the number of peripheral fractures. In efficacy and side effects are both dose- fluoride per day (20). the last study, the fluoride dosages were related. Obviously, this is a limiting fac- In an open study in 48 patients, the ef- probably too high: 60 and 90 mg on al- tor for prescribing fluoride, especially in fect of monofluorophosphate (26 mg/day ternating days. Moreover, the fluoride elderly patients with osteoporosis, in of fluoride) was compared with that of was not given in the form of enteric- whom deterioration of kidney function calcium alone (control group). After 18 coated tablets, probably leading to high may elevate blood levels of fluoride. months, BMD had increased by +7.8% (toxic) peak serum levels of fluoride In conclusion, prescribing low dose fluo- in the fluoride treated patients versus (13). ride seems to be safe, and may have posi- +3.6% in the control group (21). Thus, data on the effects of fluoride on tive effects on bone; on the contrary, pre- In another study (22) on Cs-treated pa- fractures are conflicting: both decreased scribing high dose fluoride may be as- tients, the effect on BMD of monofluor- and increased fracture rates have been sociated with side effects and may even ophosphate supplemented with calcium observed. It is suggested that high dos- increase the fracture rate. and vitamin D (group 1, n = 8 patients) ages of fluoride may (over)stimulate or of calcium and vitamin D alone (group bone formation, resulting in a large in- Effects of fluoride in the prevention 2, n = 7 patients) was compared with the crease of BMD that is not associated with and treatment of corticosteroid- BMD in 14 renal transplant patients who an increase in bone strength (or bone induced osteoporosis were not on any specific treatment (group quality); on the contrary, bone quality On the one hand, it appears to be ade- 3, historical controls). No significant may even be decreased in patients treated quate to use anti-osteoporotic drugs, e.g. changes in BMD were observed at the with high dose fluoride. bisphosphonates, in Cs-treated patients, hips, while lumbar BMD tended to rise An important side effect of fluoride in based on randomized clinical trials in in groups 1 and 2, and to decrease in patients treated with high dosages of which a bone sparing effect has been group 3, leading to a statistically signifi- fluoride are incomplete fractures, which demonstrated (16-18). On the other hand, cant difference between groups 1 and 3

S-66 Hypothalamus-pituitary-adrenocortical and -gonadal axisFluoride in RA / treatmentM. Cutolo for glucocorticoid-induced osteoporosis / W.F.EDITORIAL Lems et al.

(a) (b) Fig. 1. Effect of fluoride (NaF) compared with that of placebo on bone mineral density (BMD) of the lumbar spine (a) and hips (b) in corticosteroid-treated patients without established osteoporosis. * p < 0.05 difference in the changes in BMD between groups; ** p < 0.01 difference in the changes in BMD between groups.

(p < 0.05). without established osteoporosis, fluo- BMD of the lumbar spine in the etidro- In a randomized, double-blind study on ride prevents bone loss in the lumbar nate/fluoride group had increased +9.3% patients with Cs-treated respiratory dis- spine but does not have a positive effect (95% CI: +2.3% to +16.2%, p < 0.01), eases in which 28 patients were enrolled, on the BMD of the hips. while the BMD in the etidronate/placebo a comparison was made between the ef- We also investigated whether the admin- group was unchanged: +0.3% (95% CI: fects of 100 mg monofluorophosphate istration of low dose fluoride in addition -2.2% to +2.8%). plus calcium versus calcium alone. There to cyclical etidronate has a positive ef- For the hips, no significant changes in was a statistically significant difference fect on BMD in patients with established BMD were observed in the etidronate/ in the change in BMD of the lumbar osteoporosis during continued treatment fluoride group after 2 years: -2.5% (95% spine between the two groups after 2 with Cs (25). Forty-seven Cs-treated pa- CI: -6.8% to +1.8%), while in the eti- years (p < 0.05); the increase in BMD in tients were included in this 2-year ran- dronate/placebo group the BMD had sig- the fluoride group was 11% (23). domized, double-blind, placebo-control- nificantly decreased: -4.0% (95% CI: - We performed two randomized trials in led trial. Established osteoporosis was 6.6% to -1.4%; p < 0.01). which the effect of low dose fluoride in defined as the presence of one or more We concluded that the effect of combi- Cs-treated patients was studied (24, 25). vertebral (osteoporotic) deformities on nation treatment with fluoride and etidro- The primary endpoint in both trials was radiographs. All patients were treated nate on the BMD of the lumbar spine in the difference in the change in BMD of with cyclical etidronate, calcium and ei- Cs-treated patients with established os- the lumbar spine; neither trial was large ther sodium fluoride (25 twice daily) or teoporosis is superior to that of etidronate enough to observe a difference in the placebo. After 2 years of treatment, the alone (Fig. 3). fracture rate between patients who were treated with fluoride and those who were not. The two trials differed from each other in terms of the presence or absence of vertebral deformities at baseline. In the first study (24), which was a random- ized, double-blind placebo-controlled trial, 44 Cs-treated patients without ver- tebral fractures were enrolled in a ‘pre- vention-study’. The effects of low dose Fig. 2. Effect of fluoride (NaF)/eti- sodium fluoride (25 mg twice daily) were dronate compared with that of place- compared with that of placebo: after 2 bo/etidronate on bone mineral den- years, the difference in the change in sity (BMD) of the lumbar spine in corticosteroid-treated patients with BMD at the spine between the two groups established osteoporosis. was +5.2% (95% CI: +1.8% to +8.6%; * p < 0.05 difference in the changes p < 0.01), and at the hips +0.8% (95% in BMD between groups; ** p < 0.01 CI: -2.1% to +3.8%) (Figs. 1 and 2). difference in the changes in BMD between groups. We concluded that, in Cs-treated patients

S-67 Fluoride treatment for glucocorticoid-induced osteoporosis / W.F. Lems et al.

Conclusion RM VAN, RIJN HJM VAN, BIJLSMA JWJ: Effect J Med 1998; 292-9. Although the use of fluoride is theoreti- of high dose corticosteroid pulse therapy on 18. COHEN S, LEVY RM, KELLER M, BOLING E, cally attractive in Cs-treated patients and (markers of) bone metabolism. Ann Rheum Dis EMKEY RD, GREENWALD M: Risedronate ther- 1996; 55: 288-93. apy prevents corticosteroid-induced bone loss. although it has been shown that fluoride, 8. LAU KHW, BAYLINK DJ: Molecular mecha- Arthritis Rheum 1999; 2309-18. prescribed for the prevention and treat- nism of action of fluoride on bone cells. J Bone 19. BAYLEY TA, MULLER C, HARRISON J, BASU- ment of CIOP, increases BMD of the Min Res 1998; 1660-7. ALDO J, STURTRIDGE W, JOSSE R: The long- lumbar spine, it cannot be considered the 9. KLEEREKOPER M: The role of fluoride in the term treatment of steroid-osteoporosis with prevention of osteoporosis. Endocrin Metab fluoride. J Bone Min Res 1990; 5 (Suppl. 1): first choice for these indications because, Clinics of North America 1998; 27: 441-52. S157-161. in contrast to therapy with bisphospho- 10. REGINSTER JY, HALKIN V, HENROTIN Y, 20. GREENWALD M, BRANDLI D, SPECTOR S, nates, no positive effects on BMD of the GOSSET C: Treatment of osteoporosis: Role of SILVERMAN S, GOLDE G: Corticosteroid-in- hips and (thus far) no reduction in the bone forming agents. Osteoporosis Int 1999; duced osteoporosis: Effect of a treatment with (Suppl. 2); 91-6. slow-release sodium fluoride. Osteoporosis Int vertebral fracture rate have been shown. 11. MAMELLE N, DUSAN R, MARTIN JL et al.: 1992; 2: 303-4. Risk-benefit ratio of sodium fluoride treatment 21. RIZZOLI R, CHEVALLEY T, SLOSMAN DO, in primary vertebral osteoporosis. Lancet 1988; BONJOUR JP: Sodium monofluorophosphate References 2: 361-5. increases vertebral bone mineral density in pa- 1. ACR TASK FORCE ON OSTEOPOROSIS GUIDE- 12. RIGGS BL, SEEMAN E, HODSGON SF, TAVES tients with corticosteroid-induced osteoporo- LINES: Recommendations for the prevention DR, O’FALLON WM: Effect of the fluoride/cal- sis. Osteoporosis Int 1995; 5: 39-46. and treatment of glucocorticoid-induced oste- cium regimen on vertebral fracture occurence 22. LIPPUNER K, HALLER B, CASEZ JP, MONT- oporosis. Arthritis Rheum 1996; 39: 1791-801. in postmenopausal osteoporosis. N Engl J Med ANDON A, JAEGER P: Effect of disodium mon- 2. BIJLSMA JWJ: Prevention of glucocorticoid- 1982; 306: 446-50. ofluorophosphate, calcium and vitamin D sup- induced osteoporosis. Ann Rheum Dis 1997; 13. RIGGS BL, HODGSON SF, O’FALLON WM et plementation on BMD in patients chronically 56: 507-9. al.: Effect of fluoride treatment on the frac- treated with glucocorticosteroids: A prospec- 3. EASTELL R, REID DM, COMPSTON J et al.: A ture rate in postmenopausal women with oste- tive, randomized, double-blind study. Min UK consensus group on management of glu- oporosis. N Engl J Med 1990; 322: 802-9. Electrolyte Metabolism 1996; 22: 207-13. cocorticoid-induced osteoporosis: An update. 14. RINGE JD, KIPSHOVEN C, COSTER A, UM- 23. GUAYDIER-SOUQUIRES G, KOTZKI PO, J Intern Med 1998: 244: 271-92. BACH R: Therapy of established osteoporosis SABATIER JP, BASSSE-CATHALINAT B, LOEB 4. MANOLAGAS SC, WEINSTEIN RS: New dev- with monofluorophsphate plus calcium: dose G: In corticosteroid-treated respiratory dis- elopments in the pathogenesis and treatment related effects on bone density and fracture eases, monofluorophosphate increases lumbar of steroid-induced osteoporosis. J Bone Min rate. Osteoporosis Int 9: 171-8. bone density: A double masked study. Osteo- Res 1999; 14: 1061-6. 15. REGINSTER JY, MEURMANS L, ZEGELS B et poros Int 1996; 6 (2): 171-7. 5. MEUNIER PJ, DEMPSTER DW, EDOUARD C, al.: The effect of sodium monofluorphosphate 24. LEMS WF, JACOBS JWG, BIJLSMA JWJ, CHAPUY MC, ARLOT M, CHARDHON S: Bone plus calcium on vertebral fracture rate in post- CROONE A, HAANEN HCM, HOUBEN HHML: histomorphometry in corticosteroid-induced menopausal women with moderate osteoporo- Effect of sodiumfluoride in prevention of cor- osteoporosis in Cushing’s syndrome. Adv Exp sis. Ann Intern Med 1998; 129: 1-8. ticosteroid-induced osteoporosis. Osteoporo- Med Bio 1984; 171: 191-200. 16. ADACHI JD, BENSEN WG, BROWN J et al.: In- sis Int 1997; 7: 575-82. 6. LEMS WF, VEEN GJM VAN, GERRITS MI et al.: termittent etidronate therapy to prevent corti- 25. LEMS WF, JACOBS JWG, BIJLSMA JWJ, VEEN Effect of low dose prednisone (with calcium costeroid-induced osteoporosis. N Engl J Med GJM VAN, HOUBEN HHML, HAANEN HCM: Is and calcitriol supplementation) on calcium and 1997; 337: 382-7. addition of sodium fluoride to cyclic etidronate bone metabolism in healthy volunteers. Br J 17. SAAG KG, EMKEY R, SCHNITZLER TJ et al.: beneficial in the treatment of corticosteroid- Rheumatol 1998; 37: 23-33. Alendronate for the prevention and treatment induced osteoporosis ? Ann Rheum Dis 1997; 7. LEMS WF, GERRITS MI, JACOBS JWG, VUGT of glcocorticoid-induced osteoporisis. N Engl 56: 357-63.

S-68