The use of bisphosphonates in the treatment of osteoporosis

Bone, 13, S41-S49 (1992) Printed in the USA. All rights reserved.

8756·3282/92 $5.00

+ .00

Copyright © 1992 Pergamon Press pic

The Use of Bisphosphonates in the Treatment of Osteoporosis S. E. PAPAPOULOS, J. O. LANDMAN, O. L. M. BIJVOET, C. W. G. M. USWIK, R. VALKEMA, E. K. J. PAUWELS and P. VERMEIJ Clinical Investigation Unit of the Department of Endocrinology; Division of Nuclear Medicine of the Department of Radiology and Hospital Pharmacy, University Hospital, Leiden, the Netherlands Address for correspondence and reprints: Socrates E. Papapoulos, M.D., Department ofEndocrinology, Bldg I, C4-R, University

Hospital, Rijnsburgerweg 10,2333 AA Leiden, The Netherlands.

Abstract

(Fig. I) through which they bind strongly to hydroxyapatite crystals of the mineralized matrix, which is a prerequisite for their action on bone resorption. The rest of the structure differs among the various bisphosphonates, and it is this part of the molecule which determines their relative potency, their activityftoxicity ratio, and probably their mechanism of action. The efficacy of bisphosphonates in the treatment of conditions characterized by excessive osteoclastic bone resorption has been clearly established. Examples are the treatment of malignancy-associated hypercalcemia and Paget's disease of bone. Another extremely promising application has been the secondary prevention of skeletal complications in patients with malignancies metastasizing to bone, such as breast carcinoma. For the distinction of their clinical properties, bisphosphonates can be arbitrarily classified into three generations (Papapoulos et a1. 1989), and at least 10 different compounds are currently being clinically evaluated. These clinical developments were not, however, accompanied by equal progress in the understanding of the mechanism of action and of the pharmacokinetics of these compounds. Before, therefore, discussing the rationale of the use ofbisphosphonates in osteoporosis, some relevant pharmacological aspects are considered.

The efficacy of bisphosphonates in the treatment of conditions characterized by increased osteoclastic bone resorption has been established. Recent evidence indicates that these compounds are also effective in the treatment of patients with osteoporosis. Two main protocols have been tried. One is based on the intermittent administration of the bisphosphonate, which is expected to decrease bone resorption, and give a drug-free period during which bone formation may proceed at a normal rate, leading to a positive calcium balance. The other argues that the resetting of the equilibriumin a cyclical process is, as a rule, incomplete and continuous low-grade suppression of resorption will result in a continuing positive bone balance. Intermittent administration of the first generation bisphosphonate, etidronate, for up to three years increases trabecular bone density, stabilizes it after two years, and appears to reduce the rate of new vertebral fractures in women with postmenopausal osteoporosis. Longer follow-up studies are needed beforethis beneficial effect is unequivocally established. Continuous administration of the second-generation bisphosphonate, pamidronate, increases spinal bone density in patients with osteoporosis linearly for up to four years, and is associatedwith a lowrate of new vertebral fractures. These results need to beconfirmedin controlled studies involving more patients. There are indications that pamidronate given continuously can prevent glucocorticoid-induced bone loss. There is no informationabout the effects of bisphosphonates on non-vertebral fractures. There are limited data about the use of bisphosphonates in the prevention of postmenopausal bone loss. Extensive studies on efficacy and safety are needed before this treatment is offered as an alternative to hormone replacement therapy.

Mechanism of Action of the Bisphosphonates The precise mechanism of action of bisphosphonates on osteoclastic bone resorption has yet to be fully elucidated. There are three possible mechanisms by which bisphosphonates can affect osteoclastic resorption. First, they may interfere with the function of mature osteoclasts which have ingested bisphosphonate-covered calcified matrix. Second, they may maintain a concentration gradient at the bone surface of sufficient magnitude to directly affect cellular processes involved in the initiation of osteoclast activation. Third, they may alter properties of the bone matrix that are responsible for the final activation of the osteoclasts. In order to answer these questions, appropriate test systems were essential. From the early stages of bisphosphonate development, it became apparent that variable results could be obtained in vitro which depended largely on the test system used. Boonekamp et a!. (1986, 1987), for example, showed that the inhibitory potency of bisphosphonates on osteoclastic resorption depends on the stage of osteoclast development

Key Words: Bisphosphonates-Clodronate-Etidronate - Pamidronate - Tiludronate - Osteoporosis- Bone mass.

Introduction Bisphosphonates were initially developed as pyrophosphate analogs resistant to enzymatic hydrolysis and were found to inhibit osteoclast-mediated bone resorption (Fleisch 1983). All bisphosphonates contain a P-C-P bond 841

842

S. E. Papapoulos et al. : Bispho sphonates in osteoporosis

OH

CH 3 OH

I I I -I I I OH OH OH p-c-P

0

60

0

Etidronate

..'" ..'"

.

~ 40 ~u

c

0

-

CI

I I I I

OH

J

CI

I

ZZZ cont rol

~

="",

i 20;

0

I

Clodronate

6 Icr

10'S

10. 4

Concentrat ion (M)

I

OH 60

N\H 2

SSS str ipped bone • liver

Z?2. • .. •

~ 50

.!

~

CH 2 OH

I

I I I I I I

stripped bone alone

APO EHOP ClzMOP

... 30 0; c 20

CH 2 OH

OH

40

~u

0= P-C-p _

OH

str tpped contr ol

• APO .. EHOP • ClzMOP

I

~1

OH

P-C-P

1

1 ;! 30 0;

OH

50

0

Pamidronate

OH

Fig. 1. Chemical structures of the three most commonl y used bisphosphonates which have also bee n applied to the treatmen t of patients with osteoporosis.

in the in vitro systems. Comparison of the antiresorptive potencies of the three most commonly used bisphosphonates (etidronate, clodronate, and pamidronate) in cultures of fetal long bone explants, in which resorption depends on the activation of mature osteoclasts, showed clodronate to be about 10 times more potent than pamidronate (Fig. 2), whereas numerous clinical studies have shown the opposite to be true. Another system which has been developed in the Department of Cell Biology of the Unive rsity of Le iden (Burger et al. 1983) appeared to be more representative. In this system metacarpals (or metatarsals) of fetal mice instead of radii are used . At the time of explantation , these explants differ from the radii of the same animal in that the mineralized matrix has not yet been invaded by osteoclasts ; on ly osteoclast precursors are present and these are confined in the periosteum. These precursors still need to develop to mature resorbing cells. If the periosteum is removed and these explants are cultured for 10 days , no resorption occurs. However, when the same explants are cultured in the presence of a source of osteoclast precursors, such as fetal liver tissue, resorp tion does occur because the precursors develop into mature osteoclasts that invade and resorb the calcified matrix . When this system was used to examine the effects of different bisphosphonates on osteoclastic resorption , a different picture emerged, namely that pamidronate was found to be more potent than either etidronate or clo-

10, 5

10.4 Concenl rallo n (M)

Fig. 2. Effects of bisphosphonates on osteoclastic resorption, measured as 4$Ca relea se, in fetal rat radii (up per pa nel) and in mouse periostless metac arpals co-cultured with fetal liver (l ower panel). ( j.) EHDP = etidronate ; (_) ClzMDP = clodron ate; (e ) APD = pamidronate ; reproduced from Boonekamp et al. 1986.

dronate, as would be expected from the clinical experience (Fig. 2). To obtain further proof that the co-culture resorption system reflected accurately the relative antiresorptive molar potencies of the bisphosphonates in vivo, we used this technique prospectively in the development of a new nitrogen-containing bisphosphonate, [(3-dimethylamino-lhydroxypropylidene)-l ,l -bisphosphonate; dimethyl-APD] (Papapoulos et al. 1989). This bisphosphonate was found to be about five times more potent than pamidronate in the co-culture system. On the basis of this assessment, different doses of oral or intravenous dimethyl-APD were given to patients with Paget's disease, and suppres sion of bone resorption was examined as the fractional decrease in the excess of urinary hydroxyproline excretion as described by Harinck, Bijvoet , Blanksma et al. (1987). As shown in Fig. 3, the relative antiresorptive potencies of pamidron ate and dirnethyl-Af'D in the patients with Paget's disease were similar to those predicted in the in vitro experiments. Using the above-mentioned in vitro system , Boonekamp et al. (1986) showed that pretreatment of the osteoclast precursors with bisphosphonate did not affect their subsequent development into resorbing osteoclasts, and Lowik et al. (1988) demonstrated further that pretreatment

S43

S. E. Papapoulos et al. : Bisphosphonates in osteoporosis

products of the osteogenic cells which, in turn, are responsible for the final activation of osteoclasts.

NET % 4S Ca RELEASE

70• •~ _ ~ _ ~ ~ _ ~ _ m ~ ~ 60

=

SO

~

OSTEOClAST-DEVOID =BONE -FETAl LIVER

40

OSTEOCLAST-DEVOID

BONE ALONE

o APD • DIMETHYL- APD

3j

20

10~~~~~~_~~_W.ji~~ 10-7

10-6

10-5

CONCENTRATION (M)

10

!

o

~

o

•

>-

I ~

Cl

4

"-' u >< "-'

• •

'it.

~

0

2

5

10

20

SO

100

300

1000

BISPHOSPHONATE DOSE (mg/d)

Fig. 3. Upper panel: Dose-response relation of the inhibitory effects of pamidronate CO) and dimethyl-APD ce) on resorption of mineralized matrix in subsequent co-culture of osteoclast-devoid mouse metacarpals with an osteoclast precursor source. Lower panel: Dose-response relation of the inhibitory effects of parnidronate and dimethyl-APD on bone resorption in patients with Paget's disease. Results are expressed as time required to induce 50% inhibition of initial excess of urinary hydroxyproline. (e) IV dimethyl-APD; (0) IV pamidronate; (_) oral dimethyl-APD; CO) oral pamidronate; reproduced from Papapoulos et al. 1989.

of the explants with nitrogen-containing bisphosphonates did not inhibit the chemotaxis of osteoclast precursors towards the bone. In addition, inhibition of resorption was shown to depend on matrix-bound bisphosphonate and did not require its presence in the medium. In more recent studies, we found that the proliferation of osteoclast precursors is also not affected and that bisphosphonates exert their action at a post-mitotic level (van der Pluijm et aI. 1991); a finding which is in agreement with the studies of Hughes et al. (1989) using long-term bone marrow cultures. Though the exact mechanism responsible for bisphosphonate-inhibited osteoclastic resorption is not yet known, there is considerable evidence suggesting that at least the nitrogen-containing bisphosphonates suppress resorption by inhibiting the phenotypic transformation or the terminal differentiation of osteoclast precursors into mature osteoclasts, and not by toxicity to these cells or to their precursors. They appear to alter the properties of the mineralized matrix through interactions with cytokines or other

Pharmacodynamics of the Bisphosphonates Study of the pharmacokinetics of bisphosphonates has been limited for the following reasons: I. Development of sensitive and precise assays for their determination in biological fluids is difficult and pharmacokinetic information was mainly obtained with the use of radiolabelled compounds, precluding generally their study in humans . 2. Even when these compounds are used in animals, classical pharmacokinetic concepts and analyses are not applicable because of the long retention time of bisphosphonates in bone, which makes it difficult to obtain a pharmacokinetic steady state. 3. Bisphosphonates affect a bone surface-related process and therefore surface-bound bisphosphonate should be considered as a compartment distinct from the remainder of the drug which is taken up into the bone. At present, there are no compartmental models available which can dissociate surface-bound, and hence biologically active, from bone-bound bisphosphonate. The design, therefore, of therapeutic regimens depends to a large extent on the interpretation of pharmacodynamic data. Studies in patients with malignancy-associated hypercalcemia have shown that short-term administration of a sufficiently high dose of bisphosphonate can normalize the unopposed bone resorption (Canfield 1987, Kanis 1987, Harinck, Bijvoet, Plantingh et aI. 1987). The same is also true in patients with Paget's disease of bone, in whom increased bone resorption is coupled with increased bone formation. Administration, for example, of pamidronate (20 mgfd IV) for 10days to such patients suppresses bone resorption without any demonstrable effect on bone formation (Fig. 4). In the following months and in the absence of any further treatment, bone formation declines progressively until a new equilibrium between resorption and formation is reached. These kinetics are consistent with concepts about the rapid turnover of the osteoclasts, the life span of the osteoblasts, and the tight coupling between bone resorption and bone formation (Nijweide et al. 1986). Therefore, even a short but resorption-suppressive course with high-dose bisphosphonate can have prolonged effects on bone metabolism and may suppress bone turnover completely (or near completely) if this is not high before treatment. This effect is undesirable in patients with osteoporosis. Thus, bone cellular dynamics and pharmacodynamics in relation to dose , mode of administration, and treatment duration had to be considered. Continuous administration of pamidronate to patients with Paget's disease is also followed by a rapid suppression of bone resorption, without a concurrent effect on bone formation. In the initial phase of treatment, there is therefore a dissociation between resorption and formation in favour of the latter, resulting in an increase in calcium balance by about 8.0 mmolfd (Fig. 5). This dissociation is, however, transient due to the mechanisms which couple bone formation to bone resorption, and suppression of resorption is subsequently followed by a decrease in formation until a new equilibrium is re-established. Yet, after six to nine months of continuous administration of parnidronate the equilibrium calcium balance was found to be about 1.0 mmol/d more positive than before treatment started (Fig. 5; Frijlink et aI. 1979); this showed persistence

S. E. Papapoulos et al.: Bisphosphonates in osteoporosis

S44 . APD

20 .. g/

d·

I.V.

v.d.B.

with continuous administration of a low dose of the bisphosphonate.

200

AP

Bisphosphonatesin the Treatment of Osteoporosis

(u/U

10 0 ~

~

0 1600

OHP

1200

(IJmo I124h) 800 400 0

UCa (mm oI124 h)

Ca/ Cr

:lj

0.9 0.6

OJ

.--.

0

-2

0

2

4

6

8

The rationale for the application of bisphosphonates in the treatment of patients with osteoporosis is not straightforward for two main reasons. First, osteoporosis is not generally associated with increased osteoclastic activity; second, all available information about the cellular dynamics of bone remodelling suggest that suppression of bone resorption will be followed by suppression of bone formation to an equal extent, which will lead to a new state of low bone turnover that may be undesirable in patients with osteoporosis. The studies mentioned above, however, suggested that such an application may be possible provided that certain precautions were taken. The applicability of bisphosphonates to the treatment of patients with osteoporosis has been explored in two ways. One is based on the intermittent administration of the bisphosphonate, which is expected to decrease bone resorption and give a drug-free period during which bone formation may proceed at a normal rate, leading to a positive bone balance. The other argues that the resetting of the equilibrium in a cyclical process (formation-resorption) is, as a rule, incomplete and continuous low-grade suppression of resorption will result in a continuing positive bone balance (Bijvoet et

100

10

DAYS Fig. 4. Sequential changes of serum alkaline phosphatase activity (AP), urinary hydroxyproline (OHP), and calcium (UCa) excretion rates, and of the calcium-to-creatinine ratio in morning mines (CalCr) in a patient with Paget's disease treated with daily intravenous infusions of pamidronate (20 mg/d) for 10 days. Hatched areas represent normal ranges.

0

80

•

Alk.Pase Uhypo V

60 40 20

of the effect. The question was whether this effect could be reproduced in patients with osteoporosis and normal or low bone turnover. Valkema, Vismans, et al. (1989) showed that administration of 600 mg/d pamidronate to patients with osteoporosis led to an increase in calcium balance of 5.5 mmol/d within 15 days. The effect could therefore be reproduced. The dose of pamidronate was high, however, and if given for a long period could suppress bone metabolism to such an extent that it might make the bones more prone to fractures. The question of administering a lower dose continuously was addressed in animal studies. Reitsma et al. (1980) treated rats with daily injections of pamidronate and showed that both the rate and degree of suppression of bone resorption, measured as urinary hydroxyproline excretion, are dose-dependent. During treatment, suppression of resorption reached a plateau that was dose-dependent and did not decrease further despite continuous administration of the bisphosphonate (Fig. 6). From these studies, it could be concluded that a. the accumulation of the bisphosphonate in bone is not necessarily accompanied by a cumulative effect on bone resorption and, b. it is possible to modulate bone remodelling mildly

t_

0

5

to

20

50

100 200

days of treatment ( Log scale) Fig. 5. Changes of serum alkaline phosphatase activity, urinary hydroxyproline excretion, and calcium balance in patients with Paget's disease treated with oral pamidronate 600 mg/d; reproduced from Frijlink et al. 1979.

845

S. E. Papapoulos et al.: Bisphosphonates in osteoporosis

--*" --b- --b-- -0- -t "0'-

100

+ 60

0.16

eE 8

40

>-

-. ~ .. ~ .. ~. -.~. '1" -., -" -'

100

a

~ 0:I:

0

60 1.60

~

a E a .~

E

~

40

~

.-+-+- +-+ -1 --+--------

c

'C ::J

l'lI

~

100

<:Ii

en

a

"9 0

o,

«

60 16.0

40 j

0

5

10

15

tim9 / days Fig. 6. Rate and degree of suppression of bone resorption in groups of six rats each during treatment with daily injections of pamidronate. (D) control animals; (_) treated animals; reproduced from Reitsma et al. 1980.

al. 1990). The first protocol has been tried in controlled studies with the first-generation bisphosphonate etidronate. The second has been applied only in open, albeit long-term studies with the second-generation bisphosphonate pamidronate. Etidronate

The use of etidronate in the treatment of patients with osteoporosis is not new. As early as the 1970s, various groups tried to treat such patients by giving the bisphosphonate continuously in rather high doses (20mg/kg/d). Although slight but significant increases in calcium balance were found (Heaney & Saville 1976), the drug appeared to interfere with the mineralization of newly formed bone. Interest in its use was renewed when Anderson et al. (1984) reported increases in trabecular bone (assessed histologically) and in bone remodelling dynamics in five osteoporotic patients treated with etidronate intermittently as part of a so-called coherence or ADFR therapy (Frost 1981). Oral phosphate was used as activator of osteoclast activity (Activate), followed by two weeks of etidronate treatment (Depress), followed by a 70-day treatment-free period (Free); the cycle was then repeated (Repeat). Subsequent studies, from the same laboratory, in 37 patients with osteoporosis failed to confirm these preliminary observations, and it was concluded that this regimen resulted in a short-term improvement in trabecular bone mass with no evidence at cellular level that long-term improvements in

bone remodelling occurred (Hodsman 1989). At the same time, Pacifici et al. (1988) reported the first randomized trial with etidronate in osteoporosis. Postmenopausal women with osteoporosis were treated with a coherence therapy regimen consisting of phosphate for 3 days, followed by etidronate (400 rng/d) for 14 days, followed by 8 weeks of neither drug, plus continuous calcium carbonate therapy (lgr/d). Results were compared to those obtained in two other groups of women receiving either hormonal replacement therapy, together with calcium supplements, or calcium carbonate alone. Bone-mass measurements were made both in the axial (by QCT) and in the peripheral skeleton. At the end of the study, the bone loss at the spine was significantly greater in the ADFR-treated patients than in the hormone-calcium treated patients who did not lose any bone. These authors concluded that the regimen they used was not as effective as hormonal replacement therapy in preventing trabecular bone loss in osteoporosis, nor was it any more effective than calcium supplementation alone. This study has been criticized, in that no care was taken in the administration of calcium supplements, which should not be taken at the same time as etidronate. Calcium may bind the bisphosphonate in the intestine and reduce further its already low intestinal absorption. Recently two double-blind placebo-controlled studies of the effects of intermittent etidronate in patients with postmenopausal osteoporosis have been reported (Storm et al. 1990, Watts et al. 1990). In both studies, etidronate was given orally 400 mg/d for two weeks followed by a rest period of 13 weeks. These cycles were continued for three and two years, respectively. In the first study, vitamin D and calcium supplements were given continuously throughout the treatment period while, in the second, only calcium supplements were given during the bisphosphonate-free period. In both studies, a modest but significant increase in trabecular bone mass (DPA of the spine) was found with no change in cortical bone mass. The second study, in a separate cohort, also showed that additional administration of an activator of bone remodelling (phosphate) as part of a coherence therapy induced exactly the same responses as etidronate alone. Bone biopsies from both trials showed no adverse effects of the drug on bone histology, and it was suggested that the effect was due to reduction of the depth of the resorption cavities and a decrease in the activation frequency of new BMU s. The most important finding of both studies, which differentiates them from studies with other drugs, was the report of a significant reduction in rate of new vertebral fractures in the etidronate-treated patients in comparison to the control groups. On the basis ofthese results on efficacy and safety, etidronate has been marketed in some European countries for the treatment of established postmenopausal osteoporosis. As the drug is (or will be) generally available to members of the profession independently of their expertise in the treatment of bone diseases, it is important to critically examine some aspects of the above-mentioned studies. The populations studied differed greatly in severity of the condition and in total numbers between the two studies. In the study of Storm et al. (1990) fewer patients with much more severe osteoporosis were studied. Of the 66 patients initially recruited, 26 left the study for various reasons. It is worth mentioning that 10 patients (5 in each group) died during the period of observation. Even in this age group, this mortality should be considered high and

846

rather unusual for patients with postmenopausal osteoporosis and no underlying cause for their osteoporosis. The group appears, therefore, not to be representative of the condition. This is supported further by the documentation of a considerable loss of height, by a mean of about 2 em, and of a decrease in spinal BMC in the control. group. These are in contrast to the findings in a number of control groups in recently reported studies of the efficacy of other drugs in patients with osteoporosis and similar demographic characteristics. Furthermore, no overall. significant difference between the two groups in the rate of new fractures was found. This became apparent only after the first year of the study, when the results at one year were compared to those after 3 years. However, during this period 5 patients from the control. group as opposed to 10 patients from the etidronate group left the study, and it is not clear from the publication how the deaths were distributed between the two groups during this period. At the end of the study, X-rays were available from 20 (61 %) of 33 placebotreated patients and 18 (54%) of 33 etidronate-treated patients. The study of Watts et al. (1990) was multicentral and examined a large number of patients with very mild osteoporosis as judged by the overall spine deformities. In this study, which extended over only two years, the rates of new vertebral fractures was very low in all groups of treated patients. Comparison of the rate of new fractures in placebo-treated patients (68/1000 patient years, n = 91) against that of etidronate-treated patients (44.2/1000 patient years, n = 98) revealed no significant difference despite the numbers of patients involved. To attain a marginal level of significance all patients, including those who received phosphate in addition, had to be analyzed together. The authors analyzed further the rates of new vertebral fractures according to the bone mass of the patients at the start of treatment. When only patients with low bone mass were evaluated, a clear difference in fracture rate in favour of the etidronate-treated patients was found. However, half of the patients in all groups did not have low bone mass, and in this group fractures occurred only in etidronate-treated patients. Whether this finding bears any significance cannot be concluded inasmuch, as already mentioned, the overall fracture rates were extremely low in this study. An important consideration, finally, for both studies is that the time of assessment (up to 3 years) is too short to evaluate efficacy of treatments on fracture rates when the natural history of osteoporotic fractures is taken into account (Kanis 1984). Therefore, although these data are suggestive that intermittent administration of etidronate may have a favourable effect on new spinal fractures in patients with postmenopausal osteoporosis, they are not yet conclusive and certainly longer follow up of these groups is needed before reaching a firm conclusion about spine deformities. Clodronate

There are limited data available concerning the use of clodronate in osteoporosis. Chestnut (1984) reported that clodronate given intermittently (3 months out of 6 months) for a year to 24 patients with osteoporosis, as part of a doubleblind controlled study, led to a linear increase in total body calcium measured by neutron activation analysis. This increase continued for six months after stopping treatment. The lack of details about this treatment and the short time of observation do not allow any conclusions about the effi-

S. E. Papapoulos et al.: Bisphosphonates in osteoporosis

cacy of this regimen. In a recent report, McCloskey et al. (1990) presented preliminary data on bone mass and bone histology in 10 women with osteoporosis treated with oral phosphate and clodronate infusions, 300 mg monthly, for six months. Results were interpreted as showing suppression of osteoclastic activity within each BMU, rather than inhibition of activation of new BMUs. Results therefore similar to those obtained with continuous phosphate and intermittent calcitonin therapy. Pamidronate

As already mentioned, the second-generation bisphosphonate pamidronate has been used in open, long-term studies for treatment of patients with established osteoporosis. On the basis of the information obtained from an· imal studies and those in patients with Paget's disease, the effect of continuous oral administration of low-dose pamidronate (150 mg/d) on calcium balance in patients with osteoporosis was examined. Fourteen patients with postmenopausal or idiopathic osteoporosis were treated for one year. Mean calcium balance before treatment was -0.72 ± 0.51 mmol/d and rose significantly to 1.33 ± 0.87 mmol/d after one year (Valkerna et at 1989). On a yearly basis, this corresponds to a gain in total bone mass of 2% to 3% in a patient with an osteoporotic skeleton. This positive calcium balance is therefore of sufficient magnitude to be of interest in terms of bone mass. In an initial study, BMC of the lumbar spine was measured by DPAin 24 patients with osteoporosis treated with oral pamidronate 150 mg/d (Valkema et al. 1989). Another group of 19 patients with osteoporosis and similar clinical and densitometric findings who received the same conventional care and treatment, but no pamidronate, served as controls. In the pamidronate group, BMC increased by 6.8% ± 1.7% over a mean period of measurements of 2.2 ± 0.2 years (p < 0.0005), while no significant change occurred in the control group. The latter finding is consistent with results obtained in control groups in other studies of osteoporosis, and shows that conventional care and treatment can stabilize trabecular bone mass in patients with established osteoporosis. The increase in BMC found during pamidronate treatment was not confined to the first two years of treatment, but continued at a rate of about 3% per year. The data initially reported by Valkema et al. (1989) with additional long-term information were re-examined with a more strict and conservative statistical method that allows a precise analysis of the overall time effect of treatment (general linear model MANOVA). Results showed again a significant constant increase in lumbar BMC of 2.4% per year for at least four years of treatment. No relation was found between the changes in BMC and pre-treatment urinary hydroxyproline excretion, which implies that the increase in bone mineral content observed did not occur predominantly in patients with increased bone turnover. We also have limited data on cortical bone mass, because the SPA became available to us after the DPA. As can be seen in Fig. 7, during treatment of a small group of patients with parnidronate 150mg/d orally for two years the increase in spinal BMC did not occur at the expense of cortical bone. Long-term open studies with pamidronate have also been conducted by Belgian investigators who administered the bisphosphonate intermittently (250-300 mg/d for two months followed by a two-month drug-free period). In a recent analysis of the results in 18 patients with osteo-

S. E. Papapouios et al.: Bisphosphonates in osteoporosis

S47 1,0

115

• III

--0--

110

OPA SPA

E

e

Q)

o

..

iii >

ell III ell

-

c

CIl

105

CIl

;,g 0

0.0

~

.c 0

0,5

CIl

:::l

'D

.E

100

Cl CIl

(39) (30) (32)

(32)

(22) (16)

(13)

-0,5

J:

95 6

a

-1,0 6

12

18

24

30

-1

a

Time (months)

Fig. 7. Sequential changes of bone mineral content of the lumbar spine (DPA) and of the arm (SPA) in 14 patients with osteoporosis treated with oral pamidronate 150 mg/d.

porosis treated for at least three and up to 5 years they found an increase in BMD of the lumbar spine of 2.4% the first year, 2.5% the second year, but no further change in the subsequent three years (Devogelaer et al. 1990). They also found no significant change in cortical bone density and they concluded that the gain in bone mass with intermittent administration of pamidronate is transient, and is not due to redistribution of calcium in the skeleton. Their results are thus similar to those obtained with the use of etidronate intermittently. Whether the differences in bonemass changes observed between this and our study are due to the different mode of pamidronate administration (intermittent versus continuous) cannot be concluded from these limited data. This is, however, an issue of obvious therapeutic importance, which should be addressed in future studies. If this proves to be true, then the newer bisphosphonates, which can be administered continuously without affecting bone quality, will be preferred for the treatment of patients with osteoporosis. The effects of continuous administration of pamidronate 150 mg/d on spine deformities was examined in 40 patients with osteoporosis treated for at least two years with the bisphosphonate (Papapoulos et al. 1990). These were 32 women (all postmenopausal, mean age 65.5 ::!: 6.3 (SD) years, range 55 to 75 years) and 8 men (mean age 54.7 ± 10.2 years, range 33.5 to 67 years). All patients had at least one non-traumatic vertebral fracture (mean five fractures/patient). No patient had congenital or other metabolic bone disease or endocrine and malignant diseases; none had been on corticosteroids, bisphosphonates, or calcitonin treatment. Twenty-eight patients received calcium supplements; eight, vitamin D preparations; eight, thiazide diuretics; and five, oestrogens. All these treatments had been started at least six months before pamidronate and were continued during the study period. Before pamidronate, the mean height of the group was 166.5 ± 8.4 (SD) ern; after a mean period of treatment of 53.4 months (range 24 to 88 months) this was 166.9 ± 5.9 em. This apparent lack of change in height (Fig. 8) despite the occurrence of nine new fractures (see below) is explained by small increases in height in some of the patients due to a more straight position of the spine, because of a decrease in spinal pain and discomfort. Spine Deformity

2

3

4

5

6

7

Time (years) Fig. 8. Sequential changes of height in patients with osteoporosis treated with oral pamidronate 150 rng/d. Numbers in parentheses

represent number of patients assessed.

Index (SDI) was assessed according to Minne et al. (1988). All vertebrae from T-4 through L-5 were evalu-

ated. No significant progression in spine deformity was found (Fig. 9). Nine new vertebral fractures occurred in eight patients over a mean period of observation of 53 months. These patients had four or more vertebral fractures when they entered the study. Non-vertebral fractures occurred in three patients (in two, of the ribs and in one, of the humerus) after significant trauma. The overall rate of new vertebral fractures was 50.5(1000 patient years. The significance of this low rate is difficult to assess because of the lack of a control group. It compares favourably, however, with most data published so far in osteoporotic patients. It is also worth mentioning that all fractures occurred in patients who already had severe osteoporosis, and that the observation time (mean, more than 4 years), although not optimal, is long enough to allow preliminary conclusions about the efficacy of continuous pamidronate to be drawn. These results justify the planning of controlled studies with continuous administration of parni-

5 )(

CIl

'D

4

.: ~

..

3

CIl

2

E 0

c

CIl

c

c.

en

0

39

33

17

10

Patients (number) Fig. 9. Spine deformity index, measured according to Minne et at.

1988, of patients with osteoporosis treated with oral pamidronate 150 mg/d.

S. E. Papapoulos et al.: Bisphosphonates in osteoporosis

S48

dronate to patients with osteoporosis, and some are already under way. The long retention time and the resulting accumulation of bisphosphonates in the skeleton during long-term continuous administration has been a major concern of clinicians treating osteoporotic patients. Studies in animals treated for up to one year continuously with three nitrogen-containing bisphosphonates showed no adverse effects on the mechanical properties of bone (Ferreti, Cointry et aI. 1990, Ferreti, Mondelo et al. 1990, and Einhorn et al, 1990). With low-dose pamidronate or dimethylAPD even improvements were described. In our patients treated with pamidronate 150 rng/d continuously, serum alkaline phosphatase activity (as index of bone formation) and urinary hydroxyproline excretion (as index of bone resorption) were followed at six-month intervals. As shown in Fig. 10, after a significant decrease in both parameters after one year, no further change was observed in the following four years; a result that is in complete agreement with the animal data and shows that in man, as in experimental animals, the accumulation of the bisphosphonate in bone is not associated with a cumulative effect on bone metabolism. Additional support for this view was obtained from histological studies in some of these patients taken between two and four years of treatment, which showed no adverse effects on bone metabolism and structure (Valkema, Papapoulou et al. 1989). These findings strengthen

110...------------------.

D. ::I:

.

100

o

=>

90

the view that the surface-bound bisphosphonate is of biological importance and that, if care is taken about the administered dose, treatment will be effective and devoid of adverse effects. Prevention of Bone Loss by Bisphosphonates All the studies described so far have addressed the problem of stabilization or increase in bone mass and of protection of the skeletal integrity in patients with established osteoporosis and vertebral fractures. With regard to prevention of postmenopausal bone loss, there is at present very limited experience about the efficacy and safety of bisphosphonates in such women. Only one placebo-controlled study in postmenopausal women with tiludronate given continuously for six months has been reported (Reginster et al. 1989). This showed arrest of bone loss in the bisphosphonate-treated women, which was sustained for 6 months after stopping treatment. Extensive studies on safety and efficacy are needed before offering bisphosphonates to women for whom an established, relatively safe treatment (oestrogens) already exists. In paraplegic immobilized patients, clodronate has been shown to prevent acute bone loss at doses of 400 mgld and 1600 mg/d orally given continuously for 100 days (Minaire et al. 1987). Glucocorticoid-induced bone loss presents a major clinical problem and an efficient prophylactic treatment is urgently needed. Only one placebo-controlled study using pamidronate 150 mg/d has been published, and showed a favourable effect on both trabecular and cortical bone mass in the patients who received the bisphosphonate (Reid, King et al, 1988 & Reid, Heap et al. 1988). These are extremely encouraging results, which have the limitations of the small number of patients and the lack of data on fracture rates.

Acknowledgments: The studies from the authors' laboratory

have been supportedby Fungo (ZWO), Medigon, and NWO (grant

70 60

900-541-191); additional support was provided by Ciba-Geigy Ltd.

+---r---r---r-"""-"""T---.---.---1 -10

0

10

20

30

40

50

60

70

110 . . . - - - - - - - - - - - - - - - - - - - ,

~

e 2

=:

:! ::

100

90

c

eo

o'!

70

...o

60

+--.---,----r---r-""""T---.---.---l o 10 20 30 40 50 60 70

-10

Time

(months)

Fig. 10. Percent changes of urinary hydroxyprolineexcretion and of serum alkaline phosphatase activity in patients with osteoporosis treated with oral pamidronate 150 mg/d,

References Anderson, C.; Cape, R. D.; Crilly, R. G.; Hodsman, A. B.; Wolfe, B. M. Preliminary observations of a form of coherence therapy for osteoporosis. Calcif. Tissue Int. 36:341-343; 1984. Bijvoet, O. L. M.; Valkerna, R.; Lowik, C. w. G. M.; Papapoulos, S. E. The use of bisphosphonates in osteoporosis. DeLuca, H. F.; Mazess, R., eds, Osteoporosis: Physiological Basis, Assessment and Treatment. New York: Elsevier Science 1990; 331-338. Boonekamp, P. M.; van der Wee-Pals, L. J. A.; van Wijk-van Lennep, M. M. L.; Thesing, C. W.; Bijvoet, O. L. M. 1\vo modes of action of bisphosphonates on osteoclastic resorption of mineralized matrix. BOlle Miller. 1:27-39; 1986. Boonekamp, P. M.; L6wik, C. W. G. M.; van der Wee-Pals, L. J. A.; van Wijk-van Lennep, M. M. L.; Bijvoet, O. L. M. Enhancement of the inhibitory action of APD on the transformation of osteoclast precursors into resorbing cells after dimethylation of the amino group. Bone Miller. 2:29-42; 1987. Burger, E. H.; van der Meer, J, W. M,; van de GeveJ, J. S.; Gribnau, J. C.; Thesingh, C. W.; van Furth, R. In vitro formation of osteoclasts from long term cultures of bone marrow mononuclear phagocytes. J. Exp, Med. 156:1604-1615; 1982. Canfield, R. E" ed, Etidronate disodium: a new therapy for hypercalcemia ofmalignancy.Am.J. Med. 82:2A:I-78; 1987.

S. E. Papapoulos et al.: Bisphosphonates in osteoporosis Chestnut, C. H. Synthetic salmon calcitonin . diphcsphonates and anabolic steroids in the treatment of postmenopausal osteoporosis. Christ iansen, C. et al., eds. Osteoporosis . Glostrup Hospital. Copenhagen; 1984;549555. Devogelaar, J. P.; Nagant de Deuxchaisnes, C. Treatment of involutional osteop orosis with the bisphosphonate APD (disodiurn pamidronate); non-linear increase of lumbar bone mineral density, J. Bone Miner . Re s. 5:S250; 1990. Einhorn , T.; Peter. C. P.: Redan, G. A.; Thompson , D. D. Effect of Alendronate on mechanical properties of bone in rats and dogs. J. Bone Miner . Res. 5:S97; 1990. Ferreti , J. L.; Cointry, G.; Capozza , R. ; Montuori , E.; Roldan, E.; Perez Lloret, A. Biomechanical effects of the full range of useful doses of (3-amino-I-hydroxypropylidene). J.]·bisphosphonate (APD) on femur diaphyses and cortical bone tissue in rats . Bone Miner. 11:111-122; 1990. Ferretti, 1. L.; Mondelo, N.; Vazquez , S.; Capozza, R.; Cointry, G.; Capiglioni, R.; Bogata, C. E.; Montuori, E . ; Sanchez, T. V.; Zanchetta, 1. R. Effects of dimethyl-parnidronate on mineral density and mechanical properties of male rat femora. J. Bone Miner. Res. 5:S104; 1990. Fleisch, H. Bisphosphonatcs: mechanism of action and clinical application. Peck, W. A., ed. Bone and Mineral Re search Annual I. Amsterdam: Exerpta Medica: 1983 ;319-357. Frijlink, W. B. ; te Velde, J.; Bijvoet , O. L. M.; Heynen, G. Treatment of Paget's disease with (3-amino·l·hydroxypropylidene)-I , l-bisphosphonate CAPD). Lancet 1:799-803 ; 1979. Frost , H. M. Treatment of osteoporosis by manipulation of coherent bone cell populations . Clin. Orth , 143:227-244; 1979. Harinck, H. 1. 1.; Bijvoet, O. L. M. ; Blanksrna , H . J.; DahlinghausNienhuys, P. 1. Efficaceous management with APD in Paget's disease of bone. CUn. Orth. Rei. Res . 217:79-98; 1987. Harinck, H. 1. 1.; Bijvoet, O. L. M.; Plantingh, A. S. T.; Body, 1.; Elte, J. W. F.; Sleeboom, H. P. ; Wildring, J. ; Neyt , 1. P. The role of bone and kidney in tumor hypercalcemia and its treatment with bisphosphonate and sodium chloride . Am . J . Med. 82:1133- 1142; 1987 . Heane y, R. P.; Saville, P. D. Etidronate disodium in postmenopausal osteoporosis. CUn. Pharm acal. Thera. 20:593- 604; 1976. Hodsman , A. B. Effects of cyclical therap y for osteoporosis using an oral regimen of inorganic phosphate and sodium etidronate : a clinical and bone histomorphometric study. Bone Miner . 5:201-212; 1989. Hughes, D. E.; MacDonald , B. R.; Russell , R . G. G.; Gowen, M. Inhibition of osteoclastic-like cell formation by bisphosphonates in long-term cultures of human bone marrow. J . C line Invest , 83:1930-1935; 1989. Kanis, 1. A. Treatment of osteoporotic fracture. Lancet i:27-33; 1984. Kanis, J. A., ed. Clodronate-a new perspective in the treatment of neoplastic bone disease. Bone 8:1; 1987. Lowik, C. W. G. M.; van der Pluiim, G. ; van der Wee-Pals, L. J. A.; Bloys van Treslong-de Groot, H. ; Bijvoet , O. L. M. Migration and phenotypic transformation of osteoclast precursors into mature osteoclasts. The effeel of a bisphosphonate. J . Balle Miner. Re s. 3:185-192; 1988. McCloskey, E. V.; Beneton, M. N. C.; O'Rourke, N.; Eyres, K.; O'Doherty, D. ; Kanis, J. A. The effect of pho sphate and intermittent clodronate in postmenopausal osteoporosis. Christiansen , C. and Overgaard , K . , eds. Osteoporosis /9 90; Co penhagen: Osteopress ApS 1990:1401-1402.

S49 Minaire, P.; Depassio, J.; Berard, E .; Meunier, P. I . ; Edouard, C. ; Pilonchery, G.; Goedert, G. Effects of clodronate on immobilizat ion bone loss. Bone 8:1987. Minne, H . w. ; Leidig, G.; Wuster, Chr.; Siromachkostov , L. ; Baldauf, G. ; Bickel, R.; Sauer, P.; Lojen, M. ; Ziegler, R. A newly developed spine deior mity index (SD1) to qua ntitate vertebr al crush fractures in patients with osteoporosis. Bone Min er. 3:335-349; ]988. Nijweide, P. J.; Burger, E. H . ; Feyen, J. H. M. The cells of bone : proliferation, differentiation and hormonal regulation. Physiol, Re v. 66:855-866 ; 1986. Pacifici, R.; McMurtry. C.; Vered , 1.; Rupich, R.: Avioli, L. V. Coherence therapy does not prevent axial bone loss in osteoporotic women: a preIiminary comparat ive study . J . Clin . Endocrinol. Metab, 66:747-753 ; 1988. Papapoulos, S. E.; Hoekrnan, K. ; Lowik, C. W. G. M.; Vermeij, P.; Bij· voet O. L. M. Application of an in vitro model and a clinical protocol in the assessment of the potency of a new bisphosphonate. J. Bone Miner. Res. 4:775-781; 1989. Papapoulos, S. E.; Bijvoet, O. L. M.; Valkema, R.; Landman, J. 0.; Papapoulou, V.; Pauwels, E. K. J.; Hamdy, N. A. T; Vermeij, P. New bisphospho nates in the treatment of osteoporosis. Christiansen, C. and Overgaard, K., eds. Osteoporosis 1990. Copenhagen: Osteopress ApS, 1990:1294-1300. van der Pluijm, G.; LOwik, C. W. G. M. ; de Groot, H.; Alblas, M.; van der Wee-Pals, L. J. A.; Bijvoet, O. L. M. ; Papapoulos, S. E. Modulation of basal and PTH-stimulated osteoclastic resorption by bisphosphonates in fetal mouse bone explants . J . BOlle Min . Res, 6:1203-1210; 1991. Reginster, J. Y.; Lecart, M. P.; Dero isy, R. ; Sarlet, N.; Denis , D .; Ethgen , D. ; Collette, J.; Franchirnont, P. Prevention of postmenopausal bone loss by Tiludronate. Lancet 11:1469-1471; 1989. Reid, I. R.; King,A. R. ;A1exander, C . J. ;Ibbertson, H. K . Preventlon ateroid-induced osteoporosis with (3-amino. l- hydroxypropylidene)-I,bisphosphonate (APD). Lancet 1:143-146; 1988a. Reid, I. R.; Heap, S. W.; King, A. R.; Ibbertson, H. K. Two-year fallowup of bisphosphonate (APD) treatment in steroid osteoporosis. Lancet ii:II44; 1988b. Reitsma , P. H.; Bijvoet, O. L. M.; Verllnden-Oo ms, H,; van der Wee-Pals, L. Kinetic studies of bone and mineral metabolism during treatment with (3-amino·l·hydroxypropylidene).l,l -bisphosphonate (APD ) in rats . Calcif. Tissue Int . 32:145-157 ; 1980. Storm, T. ; Tharnsborg , G.; Steiniche , T.; Genant, H. K. ; Sorensen, O. H. Effect of intermittent cyclical etidronate therapy on bone mass and fracture rate in women with postmenopausal osteoporosis. N . Engl. J . Med. 322:1265-1271; 1990. Valkema, R.; Visrnans, F. J. F. E .; Papapoulos, S. E.; Pauwels, E. K. J.; Bijvoet, O. L. M. Maintained improvement in calcium balance and bone mineral content in patients with osteoporosis treated with the bisphosphonate APD. Bone Miner . 5:183-192; 1989. Valkema, R.; Papapoulou, V.; Pauwels, E. K. J.; Bijvoet, O. L M .; Papapoulos, S. E. No evidence of cumulative effect of low dose APD on bone remodelling during four year continuous treatment in osteoporosis . J . Bone Miner . Res. 4:5370 ; 1989. Watts, N. B.; Harris, S. T.; Genant, H . G.; Wasnich, R. D .; Miller, P. D. et al. Intermittent cyclical Etidronate treatment of postmenopausal osteoporosis. N . Engl. J. Med. 323:73-79; 1990.