Estrogen in Prevention and Treatment of Osteoporosis" ROBERT LINDSAYb," AND FELICIA COSMANCsd bDepartment of Internal Medicine, and dRegional Bone Center Helen Hayes Hospital West Hauerstraw, New York 10993 "Department of Medicine Columbia University College of Physicians & Surgeons New York, New York 10032
INTRODUCTION Almost 50 years ago Fuller Albright' first highlighted the importance of ovarian failure in the pathogenesis of osteoporosis, and the relationship between sex steroids and the skeleton has remained in the limelight since that time. This paper evaluates the role of sex steroids in the pathogenesis of osteoporosis and the place of hormone replacement in the prevention and treatment of the disorder.
PATHOPHY SIOLOGY Osteoporosis is a disorder of bone mass in which bone tissue has been lost, resulting in increased risk of fracture (FIG.1). While fracture syndromes are the direct consequence of the development of osteoporosis, other factors also alter fracture risk. Bone loss appears to be inevitable as age increases, and osteoporosis is, therefore, also inevitable for all those who live to be sufficiently old. To actually fracture, however, requires a chance occurrence, trauma, which does not occur in every individual with osteoporosis. Since the risk of falling also increases with age, those factors that cause falls are also risk factors for fracture, even though only 2% of falls result in fracture.* Falls, their cause and prevention, have yet to be adequately studied. The risk of fracture may also be increased by certain l), but qualitative bone abnormalities, including bone architecture (see FIGURE because investigation of these defects requires biopsy for bone histology, less is known about these defects than about the quantitative assessment of bone mass. Therefore these discussions will focus on bone loss, its role in fracture, and its prevention among the postmenopausal female population. Bone mass throughout life can be affected by a wide variety of factors (FIG.2). Peak bone mass (the amount of bone present in the skeleton in young adulthood) is probably determined principally by genetic factors,3 but may be modulated by nutritional and lifestyle factors, especially during the accelerated growth around puberty and perhaps also until the fourth decade. We have shown in cross-sec-
* This work was supported in part by Grant AR39191 from the U.S. Public Health Service. 326
LINDSAY & COSMAN: ESTROGEN IN OSTEOPOROSIS
327
FIGURE 1. Scanning electronmicrographs of normal (top) and osteoporotic (bottom) cancellous bone. Normal cancellous bone has trabecular plates interconnecting with small regular marrow spaces. As bone is lost, the trabeculae are penetrated and become rod-like and eventually are completely resorbed. All stages of this process can be observed. (From Dempster ef ~ 1 . Reprinted ~’ by permission.)
ANNALS NEW YORK ACADEMY OF SCIENCES
328
Men ?pause v)
Adequate Calcium Intake
2z
/-
w
- - - - - --- - -
2
0
m
10 -GENES-
0
20
-
-CALCIUM
30
--
-+
40
50
60
70
-
MECHANICAL LOADING INTAKE - - - - - - - - - -* - STRUCTURAL ERRORS
-&
FlGURE 2. Hypothetical schema of factors thought to affect bone mass in women at different stages of life. (0R. Heaney. Used by permission.)
tional studies that bone mass is dependent both on calcium nutrition and physical activity in young adult^.^ Prospective studies are ongoing to determine the amount of variability in peak bone mass that might occur with modification of these factors, but the amount seems likely to be fairly small, perhaps only 5% o r so. The development of noninvasive techniques for measurement of bone mass5 provided a breakthrough in the study of bone loss, which is essentially an asymptomatic process. Our group originally demonstrated that after only 3 years, oophorectomized women had a significantly lower bone mass than did women who had had hysterectomy with ovarian conservation.6 In a series of elegant studies, Heaney confirmed that menopause was associated with increased skeletal remodeling and changes in calcium homeo~tasis.’-~The data showed an exaggeration of the modest deficit between formation and resorption observed among premenopausal women. At the same time a reduction in calcium absorption across the intestine and increased urinary calcium loss were observed. Either of these potentially could be the cause of the bone loss, but only when increased skeletal turnover is the primary event can both occur together. In a somewhat different design to our original experiment, Richelson et a1.I0 demonstrated that 50-year-old women, 20 years after oophorectomy, had lost the same amount of bone as 70-year-old women 20 years after a natural menopause, confirming that the duration of ovarian insufficiency was a more important determinant of bone mass than age. The more recent discovery that osteoblasts appear to express classical estrogen receptors supports the concept that osteoporosis is a primary disorder of the skeleton consequent upon loss of estrogen. I ‘ , I 2 Thus the weight of evidence, including a wide variety of cross-sectional, and more recently prospective stud-
329
LINDSAY & COSMAN: ESTROGEN IN OSTEOPOROSIS
ies,I3 confirm that estrogen deficiency is the major factor that causes bone loss among the aging female population.
PREVENTION The development of noninvasive techniques for estimation of bone mass also allowed evaluation, in a prospective fashion, of estrogen intervention in prevention of postmenopausal bone loss. In a series of controlled double-blind studies, our group demonstrated that estrogen therapy prevents loss of cortical bone after oophorectomy1"17 (FIG.3). The effect is clearly seen irrespective of the time elapsed since oophorectomy before estrogen is introduced. In estrogen-treated individuals, metacarpal bone mass remains stable for at least 10 years, while loss is ongoing in the placebo-treated group. The more recent development of dualphoton absorptiometry allowed us to evaluate the effects of estrogen on the axial skeleton. After 10 years of therapy mean bone mass was 29% higher in the lumbar spine and 12% higher in the femoral neck.'* It is not clear whether this difference relates to differing rates of bone loss at these sites or relative insensitivity of bone loss in the hip to estrogen replacement. These results have been confirmed by investigators at many other centers, although primarily with studies only of 2 years' duration at most. However, overwhelming evidence supports the concept that adequate estrogen therapy, irrespective of route of administration, can retard postmenopausal acceleration of bone loss. The minimum effective dose appears to be 0.625 mg conjugated equine estrogen (Premarin) for oral therapy,17 but has not been adequately determined for other routes of administration. Some data do suggest that serum estradiol levels within the midfollicular range may be adequate to protect the skeleton.*O However, further data are required on this issue. If loss of bone tissue is indeed an important determinant of fracture risk, one would expect prevention of bone loss to reduce fracture risk. Epidemiologic data
FIGURE 3. Long-term prevention of bone loss by estrogen. The hatched area represents the placebo used (mean +SD) in a prospective controlled study in
oophorectomized women. The three lines show the mean values only for three estrogen-treated groups: treatment initiated at the time of oophorectomy (squares), 3 years (circles) or 6 years (triangles) after oophorectomy. Bone loss is prevented in all three situations. However, the earlier treatment is begun, the better the outcome, in terms of bone mass after 10 years of therapy. (From Lindsay.I9 Reprinted by permission.) I
0
I
I
2
4
I
6
I
I
I
I
I
8
10
12
14
16
Years
330
ANNALS NEW YORK ACADEMY OF SCIENCES
support this concept. Several studies have shown that estrogen therapy reduces Estimates suggest risk reduction of about 50% with at the risk of hip least 5 years of therapy, provided treatment is begun in the first few years after m e n ~ p a u s e . ’Somewhat ~ similar data can be found for fractures of the distal radius. Given that 80% of hip fractures occur among women, this would mean a reduction of about 100,000 per year if all those destined to have a hip fracture could be identified and treated at the time of menopause. It is much more difficult to find epidemiologic support for prevention of vertebral fracture, since vertebral fractures rarely result in hospital admission. However, our prospective data suggest that adequate estrogen therapy, of at least 10 years’ duration, will reduce vertebral deformity by 75-90%.16 This perhaps confirms the clinical impression that vertebral fracture risk is more determined by bone mass than trauma, while the converse may be true for the hip.
TREATMENT At present only two therapies have been approved by the FDA for treatment of patients with the established disease, that is, osteoporosis, presenting to the clinician after fracture has occurred. Both estrogen and calcitonin are expected to prevent further loss of bone mass, since both act in somewhat similar ways on the bone remodeling cycle.25 Both appear to reduce the frequency of activation of new remodeling sites, with perhaps also the additional effect of inhibition of osteoclast function. The net result is a transient increase in bone mass as remodeling cycles active at the time that therapy is initiated finish the formation period, while fewer new remodeling sites are a c t i ~ a t e d . ~Both ~ - ? ~agents are equally effective. Estrogen has the advantage of being an oral agent with well-documented effects on fracture, but its wide-ranging effects in many body systems can be a disadvantage to its use in more elderly patients. In particular, the addition of a progestogen to reduce regular menstrual bleeding, currently required for protection against chronic endometrial stimulation and consequent hyperplasia, is an unacceptable accompaniment to therapy for many patients. It is not clear whether the institution of therapy with estrogen (with or without progestogen) will alter the risk of either breast malignancy or cardiovascular disease in this older population. The bone loss that follows loss of ovarian function appears to be an intrinsic change in skeletal metabolism.’ However, it is only within the past 2 years that it has been suggested that estrogens can have direct effects on the skeleton. Two groups have presented evidence that osteoblasts or osteoblast-like cells have receptors similar to the classical estrogen receptor. l . I 2 Some data also suggest physiological responses from osteoblast-like cells upon exposure to appropriate estradiol concentrations. The changes in skeletal remodeling that occur across the menopause suggest that there is increased activation of remodeling cycles, coupled with increased rate of osteoclast activity, with both contributing to the acceleration of bone loss. The factors controlling activation frequency are not well understood, but the osteoblast population lining the resting mineral surface may well play an important role. If these cells do respond to estrogen, further messengers within bone must be present to account for the estrogen effects. Recent evidence suggests that estrogens may inhibit the production of at least two factors that are known to increase bone resorption, prostaglandins (particularly PGE2) and interleukinIn addition there is some evidence that TGF-beta and IGF-I, factors produced in bone that may stimulate bone formation, may be produced locally in greater quantities in response to
LINDSAY & COSMAN: ESTROGEN IN OSTEOPOROSIS
331
Alternative hypotheses for an indirect estrogen action on bone have been developed, but since the preliminary description was made of estrogen receptors within bone cells, these alternatives, while not completely ruled out, seem less likely. The suggestion that estrogen increases endogenous calcitonin supply has been one of the more popular explanations for estrogen action, but has been questioned as assay systems for calcitonin improve. Calcitonin levels in several systems do seem to be increased after administration of some but not all estrog e n ~In. ~uitro, ~ 17 P-estradiol increases the secretion of calcitonin from C cells.34 However, in patients with osteoporosis, it has been reported that calcitonin levels are increased, rather than decreased, suggesting that calcitonin is responding to the increased supply of calcium from bone.33Doubts about the exact physiological role of calcitonin in the pathogenesis of osteoporosis have not been dispelled as yet. However, it seems clear that calcitonin is a useful therapy inhibiting further loss of bone in patients with the established disorder. The recent availability of intranasal calcitonin has resulted in its evaluation as an alternative to estrogen for prevention of bone loss in the immediate postmenopausal period.34 Preliminary data are encouraging, but more intensive evaluation is required.
SUMMARY Estrogen has clearly been shown to decrease bone loss and frequency of osteoporotic fractures. Calcitonin has been shown in several studies to reduce the ,~~ no data are yet available demonstrating a reduction rate of bone I o s s , ~ ~ although in fracture frequency. Studies of osteoporosis intrinsically assume that prevention of further loss, or increments in bone mass, will be associated with declines in fracture recurrence. That may not always be the case. Recent controlled studies in which fluoride was used to increase bone mass in vertebral bodies resulted in no significant decline in fracture recurrence, and there was even a suggestion of increased fracture risk at some site^.'^,^^ Thus further data on calcitonin will be important, even though calcitonin is not expected to alter bone quality as does fluoride. Alternative therapies to calcitonin and estrogen are being investigated in clinical studies since both are currently limited in their use. Because calcitonin currently requires nearly daily injections, estrogen remains the principal agent available for both prevention and treatment in spite of its wide effect on multiple body systems. Bisphosphonates, given continuously or intermittently, appear to be relatively safe oral alternatives to calcitonin. The long-term effects of these agents need to be evaluated in greater detail before they can be recommended for prevention, but a role in therapy of the established disorder seems likely. The skeletal effect of bisphosphonates is also inhibition of bone remodeling and therefore prevention of further bone loss. Thus, they add nothing to the other therapeutic regimens from this perspective and as with calcitonin therapy, documentation of decreased frequency of fracture is lacking. Agents to increase bone mass are purely investigational at this time and many years may elapse before efficacy can be shown for such interventions. REFERENCES 1.
2.
ALBRIGHT, F., F. BLOOMBERG & P. H. SMITH,1940. Postmenopausal osteoporosis. Tr. Assoc. Am. Phys. 55: 298-305. CUMMINGS, S. R . , J . L. KELSEY,M. C. NEVITT& K . J . O’DOWD.1985. Epidemiology of osteoporosis and osteoporotic fractures. Epidemiol. Rev. 7: 178.
332 3. 4.
5. 6. 7. 8.
9. 10. 11.
12. 13. 14. 15.
16. 17. 18.
19. 20. 21. 22. 23. 24. 25. 26. 27.
ANNALS NEW YORK ACADEMY OF SCIENCES SOWERS, M. R., T. L. BURNS& R. B. WALLACE.1986. Familial resemblance of bone mass in adult women. Genet. Epidemiol. 3: 85-93. & R. LINDASY. 1988. Interaction of calcium nutrition and KANDERS, B., D. DEMPSTER physical activity on bone mass in young women. J . Bone Min. res. 3: 145-149. J. R. & J. SORENSEN. 1963. Measurement of bone mineral in vivo: An CAMERON, improved method. Science 142: 230-232. AITKEN,J. M., D. M. HART,J. B. ANDERSON, R. LINDSAY & D. A. SMITH. 1973. Osteoporosis after oophorectomy for nonmalignant disease. Br. Med. J. i: 325-328. HEANEY,R. P., R. R. RECKER& P. D. SAVILLE.1978. Menopausal changes in bone remodeling. J. Lab Clin Med. 92: 964-970. & P. D. SAVILLE.1978. Menopausal changes in calcium HEANEY, R. P., R. R. RECKER balance performance. J. Lab. Clin. Med. 92: 953-963. 1978. Calcium balance and calcium HEANEY,R. P., R. R. RECKER& P. D. SAVILLE. requirements in middle-aged women. Am. J. Clin. Nutr. 30: 1603-1611. RICHELSON, L. S., H. W. WAHNER,L . J. MELTON& B. L. RIGGS. 1984. Relative contributions of aging and estrogen deficiency to postmenopausal bone loss. N . Engl. J. Med. 311: 1273-1275. D. J. BENZ,et al. 1988. Estrogen binding, receptor KOMM,B. S., C. M. TERPENING, mRN A, and biologic response in osteoblast-like osteosarcoma cells. Science 241: 81-84. ERIKSEN,E. G.,D. S. COLVARD,N . J . BERG, rf a/. 1988. Evidence of estrogen receptors in normal human osteoblast-like cells. Science 241: 84-86. SLEMENDA, C., S. I,. Hui, C. LONGCOPE. & C. C. JOHNSTON.1987. Sex steroids and bone mass: A study of changes about the time of menopause. J . Clin. Invest. 80: I261- 1269. D. M. HART,E. B. MACDONALD & A. LINDSAY, R., J. M. AITKEN,J. B. ANDERSON, C. CLARK.1976. Long-term prevention of postmenopausal osteoporosis by oestrogen. Lancet i: 1038-1041. A. C. CLARK& A. KRASLINDSAY, R., D. M. HART,P. PURDIE,M. M. FERGUSON, ZEWSKI. 1978. Comparative effects of oestrogen and a progestogen on bone loss in postmenopausal women. Clin. Sci. Mol. Med. 5 4 193-195. & C. BAIRD. 1980. Prevention of spinal LINDSAY,R., D. M. HART, C. FORREST osteoporosis in oophorectomized women. Lancet ii: 1151-1 154. LINDSAY,R., D. M. HART& D. M. CLARK.1984. The minimum effective dose of estrogen for prevention of postmenopausal bone loss. Obstet. Gynecol. 63: 759-763. AL-AZZAWI,F., D. M. HART& R. LINDSAY.1987. Long-term effect of oestrogen replacement therapy on bone mass as measured by dual photon absorptiometry. Br. Med. J. 294: 1261-1262. LINDSAY, R. 1988. Sex steroids in the pathogenesis and prevention of osteoporosis. I n Osteoporosis: Etiology, Diagnosis, and Management. B. L. Riggs & L . J . Melton, 111, Eds.: 353-358. Raven Press. New York. PEACOCK, M. & P. SELBY.1986. The effect of transdermal oestrogen on bone, calcium regulating hormones and liver in postmenopausal women. Clin. Endocrinol. 25: 543547. GORDAN,G . S., J. PICCHI & B. S. ROOF. 1973 . Antifracture efficacy of long-term estrogens for osteoporosis. Tr. Assoc. Am. Phys. 86: 326-332. HUTCHINSON, T. A., J . M . POLANSKY & A. R. FEINSTEIN. 1979. Postmenopausal oestrogens protect against fracture of hip and distal radius. Lancet ii: 705-709. KREIGER, N., J. L . KELSEY& T. R. HOLFORD.1982. An epidemiological study of hip fracture in postmenopausal women. Am. J. Epidemiol. 116 141-148. 1980. Decreased risk offractures of the hip WEISS,N. S., C. L . URE& J. H . BALLARD. and lower forearm with postmenopausal use of estrogen. N . Engl. J. Med. 303: 1195-1 198. AVIOLI, L. C. The Osteoporotic Syndrome. Grune and Stratton. Orlando, FI. LINDSAY, R., D. M. HART, H. ABDALLA & F. AL-AZZAWI.1987. Inter-relationship of bone loss, its prevention and fracture expression. In Osteoporosis 11. C. Christiansen, J . S. Johansen & B. J. Riis, Eds.: 508-512. Osteopres. Copenhagen, Denmark. M. MATHEWS, W. B. NELP, K. SlSOM & GRUBER, H. E., J . L . IVEY,D. J. BAYLINK,
LINDSAY & COSMAN ESTROGEN IN OSTEOPOROSIS
28. 29. 30. 31. 32. 33. 34.
35. 36. 37.
333
C. H . CHESNUT.1984. Long-term calcitonin therapy in postmenopausal osteoporosis. Metabolism 3 3 295-303. PACIFICI, R., L. RIFAS& I. VAVED.1988. J. Bone Min. Res. 3: 204. ERNST,M., C. SCHMID, C. FRANKENFOLDT & E. R. FROESCH. 1988. Estradiol stimulation of osteoblast proliferation in uirro: mediator roles for TGFP, PGE2, JGFI. Calcif. Tiss. Res. (Suppl. 1) 42: Abstract 117. GRAY,T . K., S. MOHAN,T. A. LINKHART, M. E. WILLIAMS & D. J . BAYLINK.1988. Estrogen may mediate its effects on bone cells by signalling the observation of growth factors. J. Bone Min. Res. 3: A.552. STEVENSON, J. C., G . ABEYASEKERA, & C. J. HILLYARD.1982. Deficient calcitonin response to calcium infusion in postmenopausal osteoporosis. Lancet i: 693-695. GREENBERG, C., KIRKRWA, S. C., BOWSER,E. N., HARGIS,G., K., HENDERSON, W. T. & G. A. WILLIAMS. Effects of estradiol and progesterone on calcitonin secretion. Endocrinology 118 2594-2598. TIEGS,R. D., J. J. BODY, H. 0. WAHNER,J. BARTA,B. RIGGS& H. HEALTH,111. 1985. Calcitonin secretion in postmenopausal osteoporosis. N. Engl. J. Med. 312: 1097-1100. REGINSTER,J . Y., A. ALBERT,R. DEROISY,M. P. LECERT,M. A. FONTAINE, P. LAMBELIN & P. FRANCHIMANT. 1987. One year controlled radomized trial of prevention of early postmenopausal bone loss by intranasal calcitonin. Lancet ii: 14811483. RIGGS,B. L., S. F. HODGSON,H . W. WAHNER,J. MUHS& W. M. O'FALLON.1989. Effect of long-term fluoride therapy in bone metabolism, bone density and fracture rate in Type 1 (postmenopausal) osteoporosis. J. Bone Min. Res. 4: S418. KLEEREKOPER, M., E. PETERSEN, E. PHILLIPS,D. NELSON,S. TILLEY& A. PARFITT. 1989. Continuous sodium fluoride therapy does not reduce vertebral fracture rate in postmenopausal osteoporosis. J. Bone Min. Res. 4: 1035. DEMPSTER, D. W., E. SHANE, W. HOLBERT& R. LINDSAY. 1986. A simple method for correlative light and scanning electron microscopy of human iliac crest bone biopsies: Qualitative observations in normal and osteoporotic subjects. J . Bone Min. Res. I: 15-21.
INTRODUCTION Almost 50 years ago Fuller Albright' first highlighted the importance of ovarian failure in the pathogenesis of osteoporosis, and the relationship between sex steroids and the skeleton has remained in the limelight since that time. This paper evaluates the role of sex steroids in the pathogenesis of osteoporosis and the place of hormone replacement in the prevention and treatment of the disorder.
PATHOPHY SIOLOGY Osteoporosis is a disorder of bone mass in which bone tissue has been lost, resulting in increased risk of fracture (FIG.1). While fracture syndromes are the direct consequence of the development of osteoporosis, other factors also alter fracture risk. Bone loss appears to be inevitable as age increases, and osteoporosis is, therefore, also inevitable for all those who live to be sufficiently old. To actually fracture, however, requires a chance occurrence, trauma, which does not occur in every individual with osteoporosis. Since the risk of falling also increases with age, those factors that cause falls are also risk factors for fracture, even though only 2% of falls result in fracture.* Falls, their cause and prevention, have yet to be adequately studied. The risk of fracture may also be increased by certain l), but qualitative bone abnormalities, including bone architecture (see FIGURE because investigation of these defects requires biopsy for bone histology, less is known about these defects than about the quantitative assessment of bone mass. Therefore these discussions will focus on bone loss, its role in fracture, and its prevention among the postmenopausal female population. Bone mass throughout life can be affected by a wide variety of factors (FIG.2). Peak bone mass (the amount of bone present in the skeleton in young adulthood) is probably determined principally by genetic factors,3 but may be modulated by nutritional and lifestyle factors, especially during the accelerated growth around puberty and perhaps also until the fourth decade. We have shown in cross-sec-
* This work was supported in part by Grant AR39191 from the U.S. Public Health Service. 326
LINDSAY & COSMAN: ESTROGEN IN OSTEOPOROSIS
327
FIGURE 1. Scanning electronmicrographs of normal (top) and osteoporotic (bottom) cancellous bone. Normal cancellous bone has trabecular plates interconnecting with small regular marrow spaces. As bone is lost, the trabeculae are penetrated and become rod-like and eventually are completely resorbed. All stages of this process can be observed. (From Dempster ef ~ 1 . Reprinted ~’ by permission.)
ANNALS NEW YORK ACADEMY OF SCIENCES
328
Men ?pause v)
Adequate Calcium Intake
2z
/-
w
- - - - - --- - -
2
0
m
10 -GENES-
0
20
-
-CALCIUM
30
--
-+
40
50
60
70
-
MECHANICAL LOADING INTAKE - - - - - - - - - -* - STRUCTURAL ERRORS
-&
FlGURE 2. Hypothetical schema of factors thought to affect bone mass in women at different stages of life. (0R. Heaney. Used by permission.)
tional studies that bone mass is dependent both on calcium nutrition and physical activity in young adult^.^ Prospective studies are ongoing to determine the amount of variability in peak bone mass that might occur with modification of these factors, but the amount seems likely to be fairly small, perhaps only 5% o r so. The development of noninvasive techniques for measurement of bone mass5 provided a breakthrough in the study of bone loss, which is essentially an asymptomatic process. Our group originally demonstrated that after only 3 years, oophorectomized women had a significantly lower bone mass than did women who had had hysterectomy with ovarian conservation.6 In a series of elegant studies, Heaney confirmed that menopause was associated with increased skeletal remodeling and changes in calcium homeo~tasis.’-~The data showed an exaggeration of the modest deficit between formation and resorption observed among premenopausal women. At the same time a reduction in calcium absorption across the intestine and increased urinary calcium loss were observed. Either of these potentially could be the cause of the bone loss, but only when increased skeletal turnover is the primary event can both occur together. In a somewhat different design to our original experiment, Richelson et a1.I0 demonstrated that 50-year-old women, 20 years after oophorectomy, had lost the same amount of bone as 70-year-old women 20 years after a natural menopause, confirming that the duration of ovarian insufficiency was a more important determinant of bone mass than age. The more recent discovery that osteoblasts appear to express classical estrogen receptors supports the concept that osteoporosis is a primary disorder of the skeleton consequent upon loss of estrogen. I ‘ , I 2 Thus the weight of evidence, including a wide variety of cross-sectional, and more recently prospective stud-
329
LINDSAY & COSMAN: ESTROGEN IN OSTEOPOROSIS
ies,I3 confirm that estrogen deficiency is the major factor that causes bone loss among the aging female population.
PREVENTION The development of noninvasive techniques for estimation of bone mass also allowed evaluation, in a prospective fashion, of estrogen intervention in prevention of postmenopausal bone loss. In a series of controlled double-blind studies, our group demonstrated that estrogen therapy prevents loss of cortical bone after oophorectomy1"17 (FIG.3). The effect is clearly seen irrespective of the time elapsed since oophorectomy before estrogen is introduced. In estrogen-treated individuals, metacarpal bone mass remains stable for at least 10 years, while loss is ongoing in the placebo-treated group. The more recent development of dualphoton absorptiometry allowed us to evaluate the effects of estrogen on the axial skeleton. After 10 years of therapy mean bone mass was 29% higher in the lumbar spine and 12% higher in the femoral neck.'* It is not clear whether this difference relates to differing rates of bone loss at these sites or relative insensitivity of bone loss in the hip to estrogen replacement. These results have been confirmed by investigators at many other centers, although primarily with studies only of 2 years' duration at most. However, overwhelming evidence supports the concept that adequate estrogen therapy, irrespective of route of administration, can retard postmenopausal acceleration of bone loss. The minimum effective dose appears to be 0.625 mg conjugated equine estrogen (Premarin) for oral therapy,17 but has not been adequately determined for other routes of administration. Some data do suggest that serum estradiol levels within the midfollicular range may be adequate to protect the skeleton.*O However, further data are required on this issue. If loss of bone tissue is indeed an important determinant of fracture risk, one would expect prevention of bone loss to reduce fracture risk. Epidemiologic data
FIGURE 3. Long-term prevention of bone loss by estrogen. The hatched area represents the placebo used (mean +SD) in a prospective controlled study in
oophorectomized women. The three lines show the mean values only for three estrogen-treated groups: treatment initiated at the time of oophorectomy (squares), 3 years (circles) or 6 years (triangles) after oophorectomy. Bone loss is prevented in all three situations. However, the earlier treatment is begun, the better the outcome, in terms of bone mass after 10 years of therapy. (From Lindsay.I9 Reprinted by permission.) I
0
I
I
2
4
I
6
I
I
I
I
I
8
10
12
14
16
Years
330
ANNALS NEW YORK ACADEMY OF SCIENCES
support this concept. Several studies have shown that estrogen therapy reduces Estimates suggest risk reduction of about 50% with at the risk of hip least 5 years of therapy, provided treatment is begun in the first few years after m e n ~ p a u s e . ’Somewhat ~ similar data can be found for fractures of the distal radius. Given that 80% of hip fractures occur among women, this would mean a reduction of about 100,000 per year if all those destined to have a hip fracture could be identified and treated at the time of menopause. It is much more difficult to find epidemiologic support for prevention of vertebral fracture, since vertebral fractures rarely result in hospital admission. However, our prospective data suggest that adequate estrogen therapy, of at least 10 years’ duration, will reduce vertebral deformity by 75-90%.16 This perhaps confirms the clinical impression that vertebral fracture risk is more determined by bone mass than trauma, while the converse may be true for the hip.
TREATMENT At present only two therapies have been approved by the FDA for treatment of patients with the established disease, that is, osteoporosis, presenting to the clinician after fracture has occurred. Both estrogen and calcitonin are expected to prevent further loss of bone mass, since both act in somewhat similar ways on the bone remodeling cycle.25 Both appear to reduce the frequency of activation of new remodeling sites, with perhaps also the additional effect of inhibition of osteoclast function. The net result is a transient increase in bone mass as remodeling cycles active at the time that therapy is initiated finish the formation period, while fewer new remodeling sites are a c t i ~ a t e d . ~Both ~ - ? ~agents are equally effective. Estrogen has the advantage of being an oral agent with well-documented effects on fracture, but its wide-ranging effects in many body systems can be a disadvantage to its use in more elderly patients. In particular, the addition of a progestogen to reduce regular menstrual bleeding, currently required for protection against chronic endometrial stimulation and consequent hyperplasia, is an unacceptable accompaniment to therapy for many patients. It is not clear whether the institution of therapy with estrogen (with or without progestogen) will alter the risk of either breast malignancy or cardiovascular disease in this older population. The bone loss that follows loss of ovarian function appears to be an intrinsic change in skeletal metabolism.’ However, it is only within the past 2 years that it has been suggested that estrogens can have direct effects on the skeleton. Two groups have presented evidence that osteoblasts or osteoblast-like cells have receptors similar to the classical estrogen receptor. l . I 2 Some data also suggest physiological responses from osteoblast-like cells upon exposure to appropriate estradiol concentrations. The changes in skeletal remodeling that occur across the menopause suggest that there is increased activation of remodeling cycles, coupled with increased rate of osteoclast activity, with both contributing to the acceleration of bone loss. The factors controlling activation frequency are not well understood, but the osteoblast population lining the resting mineral surface may well play an important role. If these cells do respond to estrogen, further messengers within bone must be present to account for the estrogen effects. Recent evidence suggests that estrogens may inhibit the production of at least two factors that are known to increase bone resorption, prostaglandins (particularly PGE2) and interleukinIn addition there is some evidence that TGF-beta and IGF-I, factors produced in bone that may stimulate bone formation, may be produced locally in greater quantities in response to
LINDSAY & COSMAN: ESTROGEN IN OSTEOPOROSIS
331
Alternative hypotheses for an indirect estrogen action on bone have been developed, but since the preliminary description was made of estrogen receptors within bone cells, these alternatives, while not completely ruled out, seem less likely. The suggestion that estrogen increases endogenous calcitonin supply has been one of the more popular explanations for estrogen action, but has been questioned as assay systems for calcitonin improve. Calcitonin levels in several systems do seem to be increased after administration of some but not all estrog e n ~In. ~uitro, ~ 17 P-estradiol increases the secretion of calcitonin from C cells.34 However, in patients with osteoporosis, it has been reported that calcitonin levels are increased, rather than decreased, suggesting that calcitonin is responding to the increased supply of calcium from bone.33Doubts about the exact physiological role of calcitonin in the pathogenesis of osteoporosis have not been dispelled as yet. However, it seems clear that calcitonin is a useful therapy inhibiting further loss of bone in patients with the established disorder. The recent availability of intranasal calcitonin has resulted in its evaluation as an alternative to estrogen for prevention of bone loss in the immediate postmenopausal period.34 Preliminary data are encouraging, but more intensive evaluation is required.
SUMMARY Estrogen has clearly been shown to decrease bone loss and frequency of osteoporotic fractures. Calcitonin has been shown in several studies to reduce the ,~~ no data are yet available demonstrating a reduction rate of bone I o s s , ~ ~ although in fracture frequency. Studies of osteoporosis intrinsically assume that prevention of further loss, or increments in bone mass, will be associated with declines in fracture recurrence. That may not always be the case. Recent controlled studies in which fluoride was used to increase bone mass in vertebral bodies resulted in no significant decline in fracture recurrence, and there was even a suggestion of increased fracture risk at some site^.'^,^^ Thus further data on calcitonin will be important, even though calcitonin is not expected to alter bone quality as does fluoride. Alternative therapies to calcitonin and estrogen are being investigated in clinical studies since both are currently limited in their use. Because calcitonin currently requires nearly daily injections, estrogen remains the principal agent available for both prevention and treatment in spite of its wide effect on multiple body systems. Bisphosphonates, given continuously or intermittently, appear to be relatively safe oral alternatives to calcitonin. The long-term effects of these agents need to be evaluated in greater detail before they can be recommended for prevention, but a role in therapy of the established disorder seems likely. The skeletal effect of bisphosphonates is also inhibition of bone remodeling and therefore prevention of further bone loss. Thus, they add nothing to the other therapeutic regimens from this perspective and as with calcitonin therapy, documentation of decreased frequency of fracture is lacking. Agents to increase bone mass are purely investigational at this time and many years may elapse before efficacy can be shown for such interventions. REFERENCES 1.
2.
ALBRIGHT, F., F. BLOOMBERG & P. H. SMITH,1940. Postmenopausal osteoporosis. Tr. Assoc. Am. Phys. 55: 298-305. CUMMINGS, S. R . , J . L. KELSEY,M. C. NEVITT& K . J . O’DOWD.1985. Epidemiology of osteoporosis and osteoporotic fractures. Epidemiol. Rev. 7: 178.
332 3. 4.
5. 6. 7. 8.
9. 10. 11.
12. 13. 14. 15.
16. 17. 18.
19. 20. 21. 22. 23. 24. 25. 26. 27.
ANNALS NEW YORK ACADEMY OF SCIENCES SOWERS, M. R., T. L. BURNS& R. B. WALLACE.1986. Familial resemblance of bone mass in adult women. Genet. Epidemiol. 3: 85-93. & R. LINDASY. 1988. Interaction of calcium nutrition and KANDERS, B., D. DEMPSTER physical activity on bone mass in young women. J . Bone Min. res. 3: 145-149. J. R. & J. SORENSEN. 1963. Measurement of bone mineral in vivo: An CAMERON, improved method. Science 142: 230-232. AITKEN,J. M., D. M. HART,J. B. ANDERSON, R. LINDSAY & D. A. SMITH. 1973. Osteoporosis after oophorectomy for nonmalignant disease. Br. Med. J. i: 325-328. HEANEY,R. P., R. R. RECKER& P. D. SAVILLE.1978. Menopausal changes in bone remodeling. J. Lab Clin Med. 92: 964-970. & P. D. SAVILLE.1978. Menopausal changes in calcium HEANEY, R. P., R. R. RECKER balance performance. J. Lab. Clin. Med. 92: 953-963. 1978. Calcium balance and calcium HEANEY,R. P., R. R. RECKER& P. D. SAVILLE. requirements in middle-aged women. Am. J. Clin. Nutr. 30: 1603-1611. RICHELSON, L. S., H. W. WAHNER,L . J. MELTON& B. L. RIGGS. 1984. Relative contributions of aging and estrogen deficiency to postmenopausal bone loss. N . Engl. J. Med. 311: 1273-1275. D. J. BENZ,et al. 1988. Estrogen binding, receptor KOMM,B. S., C. M. TERPENING, mRN A, and biologic response in osteoblast-like osteosarcoma cells. Science 241: 81-84. ERIKSEN,E. G.,D. S. COLVARD,N . J . BERG, rf a/. 1988. Evidence of estrogen receptors in normal human osteoblast-like cells. Science 241: 84-86. SLEMENDA, C., S. I,. Hui, C. LONGCOPE. & C. C. JOHNSTON.1987. Sex steroids and bone mass: A study of changes about the time of menopause. J . Clin. Invest. 80: I261- 1269. D. M. HART,E. B. MACDONALD & A. LINDSAY, R., J. M. AITKEN,J. B. ANDERSON, C. CLARK.1976. Long-term prevention of postmenopausal osteoporosis by oestrogen. Lancet i: 1038-1041. A. C. CLARK& A. KRASLINDSAY, R., D. M. HART,P. PURDIE,M. M. FERGUSON, ZEWSKI. 1978. Comparative effects of oestrogen and a progestogen on bone loss in postmenopausal women. Clin. Sci. Mol. Med. 5 4 193-195. & C. BAIRD. 1980. Prevention of spinal LINDSAY,R., D. M. HART, C. FORREST osteoporosis in oophorectomized women. Lancet ii: 1151-1 154. LINDSAY,R., D. M. HART& D. M. CLARK.1984. The minimum effective dose of estrogen for prevention of postmenopausal bone loss. Obstet. Gynecol. 63: 759-763. AL-AZZAWI,F., D. M. HART& R. LINDSAY.1987. Long-term effect of oestrogen replacement therapy on bone mass as measured by dual photon absorptiometry. Br. Med. J. 294: 1261-1262. LINDSAY, R. 1988. Sex steroids in the pathogenesis and prevention of osteoporosis. I n Osteoporosis: Etiology, Diagnosis, and Management. B. L. Riggs & L . J . Melton, 111, Eds.: 353-358. Raven Press. New York. PEACOCK, M. & P. SELBY.1986. The effect of transdermal oestrogen on bone, calcium regulating hormones and liver in postmenopausal women. Clin. Endocrinol. 25: 543547. GORDAN,G . S., J. PICCHI & B. S. ROOF. 1973 . Antifracture efficacy of long-term estrogens for osteoporosis. Tr. Assoc. Am. Phys. 86: 326-332. HUTCHINSON, T. A., J . M . POLANSKY & A. R. FEINSTEIN. 1979. Postmenopausal oestrogens protect against fracture of hip and distal radius. Lancet ii: 705-709. KREIGER, N., J. L . KELSEY& T. R. HOLFORD.1982. An epidemiological study of hip fracture in postmenopausal women. Am. J. Epidemiol. 116 141-148. 1980. Decreased risk offractures of the hip WEISS,N. S., C. L . URE& J. H . BALLARD. and lower forearm with postmenopausal use of estrogen. N . Engl. J. Med. 303: 1195-1 198. AVIOLI, L. C. The Osteoporotic Syndrome. Grune and Stratton. Orlando, FI. LINDSAY, R., D. M. HART, H. ABDALLA & F. AL-AZZAWI.1987. Inter-relationship of bone loss, its prevention and fracture expression. In Osteoporosis 11. C. Christiansen, J . S. Johansen & B. J. Riis, Eds.: 508-512. Osteopres. Copenhagen, Denmark. M. MATHEWS, W. B. NELP, K. SlSOM & GRUBER, H. E., J . L . IVEY,D. J. BAYLINK,
LINDSAY & COSMAN ESTROGEN IN OSTEOPOROSIS
28. 29. 30. 31. 32. 33. 34.
35. 36. 37.
333
C. H . CHESNUT.1984. Long-term calcitonin therapy in postmenopausal osteoporosis. Metabolism 3 3 295-303. PACIFICI, R., L. RIFAS& I. VAVED.1988. J. Bone Min. Res. 3: 204. ERNST,M., C. SCHMID, C. FRANKENFOLDT & E. R. FROESCH. 1988. Estradiol stimulation of osteoblast proliferation in uirro: mediator roles for TGFP, PGE2, JGFI. Calcif. Tiss. Res. (Suppl. 1) 42: Abstract 117. GRAY,T . K., S. MOHAN,T. A. LINKHART, M. E. WILLIAMS & D. J . BAYLINK.1988. Estrogen may mediate its effects on bone cells by signalling the observation of growth factors. J. Bone Min. Res. 3: A.552. STEVENSON, J. C., G . ABEYASEKERA, & C. J. HILLYARD.1982. Deficient calcitonin response to calcium infusion in postmenopausal osteoporosis. Lancet i: 693-695. GREENBERG, C., KIRKRWA, S. C., BOWSER,E. N., HARGIS,G., K., HENDERSON, W. T. & G. A. WILLIAMS. Effects of estradiol and progesterone on calcitonin secretion. Endocrinology 118 2594-2598. TIEGS,R. D., J. J. BODY, H. 0. WAHNER,J. BARTA,B. RIGGS& H. HEALTH,111. 1985. Calcitonin secretion in postmenopausal osteoporosis. N. Engl. J. Med. 312: 1097-1100. REGINSTER,J . Y., A. ALBERT,R. DEROISY,M. P. LECERT,M. A. FONTAINE, P. LAMBELIN & P. FRANCHIMANT. 1987. One year controlled radomized trial of prevention of early postmenopausal bone loss by intranasal calcitonin. Lancet ii: 14811483. RIGGS,B. L., S. F. HODGSON,H . W. WAHNER,J. MUHS& W. M. O'FALLON.1989. Effect of long-term fluoride therapy in bone metabolism, bone density and fracture rate in Type 1 (postmenopausal) osteoporosis. J. Bone Min. Res. 4: S418. KLEEREKOPER, M., E. PETERSEN, E. PHILLIPS,D. NELSON,S. TILLEY& A. PARFITT. 1989. Continuous sodium fluoride therapy does not reduce vertebral fracture rate in postmenopausal osteoporosis. J. Bone Min. Res. 4: 1035. DEMPSTER, D. W., E. SHANE, W. HOLBERT& R. LINDSAY. 1986. A simple method for correlative light and scanning electron microscopy of human iliac crest bone biopsies: Qualitative observations in normal and osteoporotic subjects. J . Bone Min. Res. I: 15-21.
