Short-Term Intravenous Bisphosphonates in Prevention of Postmenopausal Bone Loss



This study was performed to test the efficacy of short-term intravenous clodronate and etidronate in the prevention of postmenopausal bone loss. Healthy postmenopausal women, exhibiting a decreasing trend in bone mineral density, were randomized to five groups (clodronate at doses of 150, 300, and 600 mg; etidronate at a dose of 300 mg; and a placebo group) of 21–22 subjects. The drugs were administered intravenously three times with 1-week intervals, followed by regular evaluation for up to 24 months. During the first year, 300 mg of clodronate retarded bone loss significantly in the lumbar spine and femoral neck, where significant protection still persisted after 24 months. Other doses of clodronate (150 and 600 mg) were not bone protective. Etidronate (300 mg) retarded bone loss significantly in the lumbar spine up to 24 months, relative to placebo. Serum concentrations of procollagen I carboxy-terminal propeptide and urinary Ca2+ and hydroxyproline excretion decreased in all bisphosphonate groups during the first month after treatment, but the values returned later toward baseline. In the etidronate-group, serum osteocalcin concentrations also decreased significantly during the first 3 months of the study. Otherwise, no uniform serum responses to bisphosphonate-treatment were detected in circulating markers of bone formation, alkaline phosphatase, or osteocalcin. No significant differences in the serum concentrations of cross-linked carboxy-terminal telopeptide of type I collagen were detected between the groups. Patient acceptance of both bisphosphonates was excellent, and no drug-related adverse side effects were detected. These results suggest that infrequently repeated intravenous treatment with bisphosphonates may effectively counteract postmenopausal bone loss.


Bisphosphonates (BPs), such as clodronate and etidronate, are nonhydrolyzable, pyrophosphate analogs and inhibitors of osteoclastic bone resorption that have an established role in the treatment of metabolic and metastatic bone diseases.1,2 Pharmacokinetic properties of BPs include poor absorption from the gastrointestinal canal, rapid clearance from the circulation, and accumulation in the skeleton, where they remain for prolonged periods.3,4

The exact molecular mechanisms of action of various BPs, which are likely to be different, are not known.5–7 Convincing evidence from experiments in vitro favors the view that BPs inhibit the action of mature osteoclasts, which resorb bisphosphonate-covered bone.5–7 Recently, an indirect action via osteoblasts has also been suggested.8

Bisphosphonate administration is now also considered an attractive alternative in the prevention and treatment of osteoporosis.9–15 Clodronate has long been employed in the treatment of high turnover bone diseases, but it has not been extensively used in the treatment of primary osteoporosis.16–20 Clodronate is generally well tolerated and its administration is not associated with the acute phase reaction or mineralization defects detected with amino-containing BPs20–22 or with high doses of etidronate, respectively.

BPs are typically administered intravenously, when a rapid reduction of serum calcium concentrations or pain relief in tumoral bone disease is needed.23–25 This strategy, which bypasses the typical gastrointestinal adverse effects of oral BPs, has recently been extended to the treatment of other metabolic bone diseases as well.14,26,27 We carried out a prospective, randomized, double-blind, placebo-controlled study in 107 postmenopausal women to test the effects of short-term, intravenous clodronate and etidronate on the preservation of bone mineral density (BMD). In addition, the dose-response relationship between clodronate and etidronate treatment, changes in biochemical markers of bone turnover, and study subject compliance to the therapy were investigated over a 24-month period.



Two hundred and one women, aged 50–55 years, randomly selected from the city records of Oulu, were recruited for prestudy measurements of BMD of the lumbar spine. This measurement was carried out twice, with a 6-month interval. The study subjects (n = 107) were chosen on the basis of a decrease in BMD of the lumbar spine, as detected by dual-energy X-ray absorptiometry (DEXA) (Lunar DPX, Lunar Radiation Corporation, Madison, WI, U.S.A.) during the prestudy follow-up period. In the study population, the mean BMDs decreased from 1.053 ± 0.142 to 1.031 ± 0.143 (mean ± SD) during the prestudy period. All study subjects were generally healthy women within 0.5 to 4 years after menopause (defined as the cessation of menstruation). The exclusion criteria were: (1) malignancy of any kind within the previous 5 years; (2) estrogen treatment, currently or in the preceding 2 months; (3) medication with steroids or thiazide; (4) elevated liver function test results or elevated serum creatinine concentrations; and (5) known adverse reaction to tetracycline. None of the studied women had been previously treated with calcitonin or bisphosphonates either.

Study design

The study was performed as a randomized, placebo-controlled study for 12 months. Follow-up during the second year was done openly. The study subjects were assigned to one of five treatment groups using random number tables: (A) 150 mg, (B) 300 mg and (C) 600 mg clodronate, (D) 300 mg etidronate, (E) and the placebo group. Clodronate (Bonefos, Leiras Oy, Turku, Finland) and etidronate (Didronel, Norwich-Eaton Pharmaceuticals, Norwich, NY, U.S.A.) were dissolved in 500 ml of 0.9% sodium chloride and administered intravenously (iv) over 3 h, once a week, during a 3-week period on an outpatient basis. Thus, our short-term treatment regimen consisted of three iv administrations of each dose. No further treatments were performed during the study, which lasted 24 months.

Eight subjects in each group were randomly selected for iliac crest biopsy. These participants took tetracycline hydrochloride (3× 500 mg daily) for 2 days at 10-day intervals, prior to bone biopsy at the 6-month visit. All study subjects had a complete physical examination including laboratory tests before entering the study. Informed, written consent was obtained from all the participants. The study was run according to the Declaration of Helsinki and was accepted by the local ethical committee.

Bone density measurements

BMD measurement of the lumbar spine (L2–L4) and the proximal femur (neck, Ward's triangle, and trochanter region) were measured at the beginning of the study and at 6-, 12-, and 24-month visits by DEXA. The response to treatment was calculated as the percent change from baseline.

Bone histomorphometry

Transiliac bone specimens were processed undecalcified for static and dynamic histomorphometric evaluation as described earlier.28 Analysis of bone samples were carried out using the Osteoplan II-program (Kontron, Munich, Germany) according to previously published methods.29 A minimum of 50 optical fields per sample were evaluated using 200× magnification. Parameters examined included: bone volume/total volume (BV/TV, %), osteoid volume/bone volume (OV/BV, %), osteoid surface/bone surface (OS/BS, %), eroded surface/bone surface (ES/BS), and mineral apposition rate (MAR; μm/day).30

Markers of bone turnover

Concentrations of fasting serum procollagen I carboxy-terminal propeptide (PICP) and cross-linked carboxy-terminal telopeptide of type I collagen (ICTP) were measured by radioimmunoassays (RIAs) (Orion Diagnostica, Espoo, Finland). Concentrations of fasting serum osteocalcin were measured by RIA (CIS BioInternational, Gifs/Yvette, France) and those of total alkaline phosphatase (ALP) by standard methods. All measurements, except those for osteocalcin, were carried out at onset and at 1, 3, 6, and 12 months.

Serum and urine biochemistry

Serum 25-hydroxyvitamin D3 (25(OH)D3) concentrations were determined by RIA (Incstar, Stillwater, MN, U.S.A.) at the onset of the study. The fasting serum calcium concentration, phosphate, and albumin were measured by routine laboratory methods. Serum intact parathyroid hormone (PTH) was measured by immunoradiometric assay (IRMA) (Incstar). Fasting urinary calcium (U-Ca) and urinary hydroxyproline (U-HOP) were adjusted to urine creatinine concentrations (U-Crea) and expressed as U-Ca/U-Crea and U-HOP/U-Crea. Safety measurements included standard hematology parameters (red blood cells [RBCs], white blood cells [WBCs], thrombocytes, erythrocyte sedimentation rate), serum alanine aminotransferase (ALT), and creatinine (Crea) and were analyzed by automated methods. All the above-mentioned measurements were assessed at admission and at 1, 3, 6, and 12 months after the onset of the study.


Baseline testing for all background variables, BMD, and biochemical and hematological variables were carried out by one-way analysis of variance (ANOVA). At later time points, BMD measurements and all serum parameters as well as hematological variables were analyzed by repeated measures by ANOVA. In case of significant treatment group and time interaction, the pairwise comparisons were performed with linear contrasts. No adjustment for the comparisons was made. Bone biopsy variables and relative urine variables, including those at baseline, were analyzed by using the Kruskal-Wallis test. Change within groups was tested by using Friedman's test (urine parameters). A significance level of 0.05 was used.



At the onset of the study, the participants in the various treatment groups were comparable demographically and as regards the biochemical variables. Mean serum concentrations of PTH, albumin, creatinine, osteocalcin, vitamin 25(OH)D3, PICP, ICTP, Ca, P, and ALP, as well as urinary parameters (U-Ca/U-Crea and U-HOP/U-Crea) were within normal ranges, and there were no significant differences between the treatment groups. No significant differences in the BMDs of lumbar spine (L2–L4), trochanter, Ward's triangle, or femoral neck were detected between the various treatment groups at the onset of the study (Tables 1 and 2).

Table Table 1. Subject Characteristics and BMD Values of the 107 Study Subjects at Onset
original image
Table Table 2. Serum and Urine Biochemistry at Onset and Various Time Points Duringthe Study
original image

Changes in BMD

The changes of BMD in the various treatment groups are shown in Fig. 1. At 12- and 24-month visits, the percentage changes (mean ± SD) in BMD of vertebrae L2–L4 in the placebo group were −1.9 ± 4.2 and −2.5 ± 4.6%, respectively. The corresponding changes in the other groups were: −1.2 ± 2.9 and −2.7 ± 3.7 in group A (clodronate 150 mg), 0.7 ± 3.0 and −1.7 ± 3.9 in group B (clodronate 300 mg), −1.2 ± 3.2 and −3.1 ± 3.6 in group C (clodronate 600 mg), and 0.9 ± 2.8 and −0.4 ± 3.2 in group D (etidronate 300 mg) (Fig. 1A). During the first year of the study, the BMD values of the lumbar spine had decreased significantly less in groups B and D, as compared with those in the placebo group (p = 0.022 and p = 0.002, groups B and D vs. placebo, respectively). In group D, significant protection against bone loss in the lumbar spine was still detectable after 24 months (p = 0.009 group D vs. placebo).

Figure FIG. 1.

Mean changes in BMD during the 24-month follow-up period. Data are expressed as mean % change from baseline. *p < 0.05 versus placebo.

In group E (placebo), the percentage changes in the mean BMD values of the femoral neck were −1.3 ± 2.8 and −3.2 ± 3.5% (mean ± SD) at 12- and 24-month visits, respectively. The corresponding changes in the treatment groups were: −1.0 ± 3.6 and −3.8 ± 3.8 (group A, clodronate, 150 mg), 0.7 ± 3.0 and −1.2 ± 3.9 (group B, clodronate, 300 mg), 0.7 ± 2.9 and −1.9 ± 3.3 (group C, clodronate, 600 mg), and −0.6 ± 3.4 and −1.0 ± 4.4 (group D, etidronate, 300 mg) (Fig. 1B). When compared with changes in the placebo group, BMD values of the femoral neck in group B had declined significantly less at various time points (p = 0.013, p = 0.026, and p = 0.026, BMD changes in group B vs. placebo at 6, 12, and 24 months, respectively). At 12 months, there was also a marginally significant difference in the changes of BMD values between groups C (clodronate, 600 mg) and E (placebo): the BMD of the femoral neck had decreased less in group C (p = 0.051). During the first 6 months of the study, the decrease of the femoral neck BMD was significantly less in group D when compared with the placebo group (p = 0.03 group D vs. placebo at 6 months). The mean BMD of femoral neck continued to decrease less in group D throughout the study, but at later time points, these differences were not significantly different from those of the placebo group.

In the trochanter region and in Ward's triangle, there were no statistically different changes of BMD between the various treatment groups during the follow up (Figs. 1C and 1D). In each study group, there was an overall, significant decrease of BMD values during the 24-month study period at all measurement sites.

Bone histomorphometry

Iliac crest bone biopsy specimens were obtained from 34 of the 107 participants at the 6-month visit. Two specimens were disqualified as a result of an inadequate sample, and a total of 32 bone biopsies were subjected to histomorphometric analysis. There were no significant differences in any of the parameters (bone volume [BV/TV], osteoid volume [OV/BV], osteoid surface [OS/BS], eroded surface [ES/BS], mineral apposition rate [MAR]) studied (Table 3).

Table Table 3. Bone Histomorphometry Variables at 6 Months
original image

Markers of bone metabolism

A summary of serum and urine biochemistry is presented in Table 2. Serum concentrations of PICP (S-PICP) remained constant in the placebo group but decreased in all the treatment groups during the first 1–3 months of treatment. These decreases later returned toward baseline values. The differences reached significance with clodronate at doses of 150 and 600 mg relative to the placebo group (p = 0.0075 and p = 0.0065 vs. placebo, respectively). Also in group D (etidronate 300 mg), S-PICP concentrations decreased significantly up to 3 months after treatment, relative to placebo (p = 0.03 and 0.0012 group D vs. placebo at 1 and 3 months, respectively).

There were no significant changes in the serum ALP concentrations between the various study groups during the 12-month study period. During this time period, the mean serum ALP concentrations increased in all study groups.

Etidronate treatment induced a significant reduction in serum osteocalcin levels during the first 3 months of the study, relative to placebo (group D vs. placebo p = 0.0007 at 3 months). This decrease in serum osteocalcin levels in group D was also significantly different from that detected in group C (clodronate 600 mg); in the etidronate group, the decrease of serum osteocalcin concentrations was more pronounced (300 mg etidronate vs. 600 mg clodronate, p = 0.035). No other significantly different changes were detected in serum osteocalcin levels.

Changes in serum ICTP concentrations were not significantly different between the various groups. Serum concentrations of 25(OH)D3 were within the normal range in all study groups at the starting point, and were not repeatedly measured. Serum calcium concentrations increased in group C (clodronate 600 mg) during the first 3 months of the study. Relative to placebo, this increase, which later declined, was statistically but not clinically significant (p = 0.017, group C vs. placebo group). In all groups, the serum calcium values were comparable at baseline and at the 12-month visit.

There were significant changes in the mean serum inorganic phosphate (S-Pi) responses to treatment in groups A (clodronate, 150 mg), B (clodronate, 300 mg), and D (etidronate, 300 mg). In the etidronate group, S-Pi concentrations increased during the first month, then decreased up to 3 months. In the clodronate groups, the changes were opposite (p = 0.0021 and p = 0.013, 150 and 300 mg clodronate vs. etidronate, respectively). Compared with those in the placebo group, these values did not reach significance.

Mean serum PTH concentrations were comparable and within the normal range at the onset and throughout the study in the various treatment groups. In the whole study cohort, a significant overall decline in mean PTH concentrations was detected at 3 months, but baseline PTH concentrations were attained by 12 months.

Urine biochemistry

In all the BP-treated groups, a decreasing trend of U-Ca/U-Crea and U-OH/U-Crea excretion was observed at 1 month. In the placebo group, U-Ca/U-Crea and U-OH/U-Crea slightly increased during this period. None of these changes in urine biochemistry were, however, significantly different between the various study groups, and 12-month values were comparable to those at baseline.

Hematological effects and safety parameters

Safety indices remained within the normal range throughout the study. There were no significant differences in blood hemoglobin, hematocrit, erythrocyte, platelet, or leukocyte values between the various study groups at the onset or during the course of the study. Liver and kidney functions were followed by monitoring serum ALT, albumin, and creatinine concentrations. These variables remained within their normal range throughout the follow-up period in all study groups (data not shown).

Adverse effects and discontinuations

All 107 participants continued the study for 12 months. However, three participants discontinued at this point. Two women started estrogen treatment as a result of difficult menopausal symptoms, and one participant was diagnosed as having endometrial carcinoma. These three women were excluded from the study at this time point. The study was carried on “open” for another 12 months, and the total follow-up period was thus 24 months. In general, the intravenous administration of bisphosphonates was well tolerated and caused no side effects directly related to the drugs.


The results of this study showed that one short treatment period consisting of three intravenous doses of clodronate (300 mg × 3) slightly but significantly retarded bone loss for up to 1 year in the lumbar spine and up to 2 years in the femoral neck, the skeletal sites most often associated with an osteoporotic fracture.13 This was also the first study to show that intravenously administered etidronate (300 mg × 3) provides protection against bone loss in the lumbar spine for up to 2 years. No signs of osteomalacia, which may complicate long-lasting etidronate treatment,1,4 were detected with the dosage of etidronate used in our study.

The bone-sparing effects of clodronate were dependent on the dosage used, but no clear dose-response effect was observed. The best results were obtained with 300 mg of clodronate, repeatedly given in 1-week intervals. These findings parallel those of a recent study by Liberman et al.,15 where osteoporotic women were treated with various doses of oral alendronate. In that study also, doubling of the optimal dose did not provide further protection against bone loss.

In all BP-treated groups, serum concentrations of PICP decreased during the first 1–3 months of the treatment, indicative of a decrease in bone turnover. Etidronate also induced a significant decrease in serum osteocalcin concentrations during the first 3 months of the study. Also, urinary calcium and hydroxyproline excretion decreased in all but the placebo group during the first month after treatment, as typically detected in response to antiresorptive therapy.26,27,31 In contrast to the results of Filipponi et al.,14 serum concentrations of ICTP, a marker of bone resorption,32 were not significantly different among the study groups.

Our treatment strategy with 300 mg of clodronate was not as bone-protective as oral, cyclical clodronate treatment as presented by Giannini et al.19 Our results are, however, qualitatively similar to those obtained by Filipponi et al.,14 who administered 200 mg of iv clodronate for 24 months at 1-month intervals. At that dose, no bone loss from the lumbar spine took place during the 2-year follow-up period. The apparent differences in efficacy may be related not only to different treatment regimes but also to the postmenopausal status of the study population. Timing of intravenous BP treatment may be of importance: better results are obtained the sooner after menopause treatment is initiated. In addition, with short-term treatment, BPs deposited on a resorption surface may become buried by newly formed matrix and lose biological activity.4,6 Hence, for longer lasting results, a more frequently repeated dosage is likely to be more beneficial. Regularly repeated parenteral drug administration may, however, result in problems with patient compliance in the long run. Based on the trend toward decreasing concentrations of markers of bone turnover for up to 3 months detected here, an optimal dosage might be intravenous infusion repeated four times per year.

Several parameters of bone histomorphometry did not reveal any significant differences between the various treatment groups. To gain more insight into the mechanisms of action of clodronate and etidronate in vivo, an earlier sampling time point after administration of the drugs might be more informative.

In conclusion, intravenously administered clodronate at a dose of 300 mg, given three times in 1-week intervals, may counteract postmenopausal bone loss for at least 1 year in the lumbar spine and up to 2 years in the femoral neck. Also, intravenously given etidronate at the dose of 300 mg provided protection against bone loss for up to 2 years in the lumbar spine. For longer lasting results, a more frequently repeated dose is likely to be more effective.


We thank M.Sc. Eliisa Loyttyniemi for expert statistical assistance.