The efficacy of oral clodronate 800 mg daily to reduce vertebral fractures was studied in 593 women with postmenopausal or secondary osteoporosis. The incidence of vertebral fractures was significantly reduced by 46%. The effect was not modified by the underlying cause of osteoporosis or other baseline factors including bone mineral density, QUS, weight, and smoking.
Introduction: This study aimed to determine if the bisphosphonate, clodronate (Bonefos), reduced the incidence of vertebral fractures in osteoporotic women.
Materials and Methods: Women fulfilling the WHO criteria for osteoporosis at the lumbar spine (T-score ≤ −2.5) and/or with at least one prevalent vertebral fracture were recruited to a 3-year double-blind, placebo-controlled study. A total of 593 patients were randomized to two strata comprised of women with postmenopausal osteoporosis (I, n = 483) and secondary osteoporosis (II, n = 110). They received either clodronate 800 mg daily orally (n = 292) or an identical placebo (n = 301). All patients received a calcium supplement of 500 mg daily. BMD was measured at 6, 12, 24, and 36 months, and lateral spine radiographs were obtained at baseline and annually thereafter for vertebral morphometry.
Results: Treatment with clodronate was associated with a significant increase in mean spine BMD over 3 years (percent change from baseline, 4.35 ± 6.34% versus 0.64 ± 6.02% in the placebo group, p < 0.0001). At the hip, clodronate maintained total BMD, whereas a significant decrease was observed in the placebo group (percent change from baseline 0.70 ± 5.67% versus −3.03 ± 6.32% in the placebo group, p < 0.0001). The changes at the spine and hip were similar in both strata. Incident vertebral fractures at 3 years were observed in 63 women in the placebo group and 33 patients receiving clodronate (relative risk, 0.54; 95% CI, 0.37–0.80; p = 0.001). Clodronate significantly reduced vertebral fracture risk in both strata and in women with or without prior vertebral fracture at baseline. Nonvertebral osteoporosis-associated fractures occurred in 21 women in the placebo group and in 14 women treated with clodronate. Treatment was well tolerated, with no significant difference in adverse event rates, including esophagitis, during clodronate treatment.
Conclusion: We conclude that clodronate 800 mg daily is a safe and effective treatment to reduce fracture risk in women with osteoporosis, regardless of causation.
Vertebral fractures are a frequent complication of established osteoporosis. Their clinical consequences include back pain, height loss, and a significantly increased future risk of fracture at vertebral and nonvertebral sites.(1,2) A reduction in the incidence of vertebral fractures is widely accepted as a demonstration of efficacy in the treatment of osteoporosis. Several double-blind placebo-controlled studies have shown that the bisphosphonate clodronate significantly decreases the incidence of both vertebral and nonvertebral fractures in patients with osteolysis caused by breast cancer and myeloma.(3,4) In a double-blind placebo-controlled study of women with primary breast cancer undergoing systemic endocrine and chemotherapy, oral clodronate 1600 mg daily prevented bone loss and decreased the incidence of skeletal metastases.(5,6) Similar results on BMD have been reported in two randomized, controlled studies in women with primary breast cancer.(7,8)
Controlled studies have also demonstrated beneficial effects of clodronate, administered orally or parenterally, on BMD at the spine and hip in patients with osteoporosis. In open studies of postmenopausal osteoporosis, effects have been demonstrated with clodronate doses ranging from 400 mg daily (orally) during each third month of treatment(9,10) to 200 mg (intravenously) every 3 weeks.(11) A recent placebo-controlled study has demonstrated that, when given orally, a dose of clodronate 800 mg daily is superior to that of 400 mg daily.(12)
Using a retrospective control population, a 45% reduction in vertebral fracture incidence was reported in an open study by Filipponi et al.(11) over a 6-year exposure to treatment with clodronate in women with postmenopausal osteoporosis. However, there have been no double-blind placebo-controlled prospective studies to address the efficacy of clodronate to reduce vertebral fracture risk in osteoporosis, which was the major aim of this study. An additional aim was to determine the extent that risk factors, identified at the start of treatment, might influence the response to clodronate. This is an important component of the study. For example, there is evidence that quantitative ultrasound at the heel can predict fracture risk, but it has not been demonstrated that patients with low values for ultrasound would be sensitive to the effects of antiresorptive therapy.
MATERIALS AND METHODS
The study included women with postmenopausal or secondary osteoporosis recruited from women referred to each study center for investigation and treatment of probable osteoporosis. The patients' characteristics and the efficacy of clodronate to reduce vertebral fracture incidence over the first year of treatment have been published previously.(13) Briefly, all women had densitometrically proven vertebral osteoporosis (spine T-score ≤ −2.5, using the reference ranges provided by the manufacturers) and/or had at least one prevalent vertebral fracture at entry. They were recruited at five centers in the United Kingdom, with the majority (88%) being enrolled at the Universities of Sheffield and Manchester. The protocol was approved by the Local Research Ethics Committee of each participating center and was conducted according to Good Clinical Practice.
The randomization was carried out in two strata comprised of women with postmenopausal osteoporosis or (premenopausal or postmenopausal) secondary osteoporosis. The procedure comprised a block (size 8) randomization by center and stratum. The sample size calculation was confined to the numbers needed in stratum 1 (postmenopausal osteoporosis) to detect a 40% decrease in the proportion of patients sustaining new vertebral fractures, assuming an annual incidence of 5.5% in placebo-treated patients (α = 0.05 and β = 0.8). The required sample size was 488, assuming a dropout rate of 20%. A planned minimum of 100 patients was to be recruited to stratum II.
Patients excluded from the study included those receiving treatment for a malignancy; those currently taking medication likely to influence skeletal metabolism or the interpretation of results (i.e., >500 mg daily calcium supplements, estrogens, progestogens, calcitonin, anabolic steroids, and bisphosphonates); those who had received bisphosphonates up to 1 year before enrollment; and those known to have malabsorption. After randomization, the participants received the study medication, which was comprised of either clodronate as a single daily oral dose of two 400-mg capsules (Bonefos; Leiras Oy, Helsinki, Finland) or an identical placebo. All patients also received a calcium supplement of 500 mg daily. Participants were instructed to take the study medication on an empty stomach at least 2 h before calcium-containing food or liquids. If desired by the patients, the capsules could be taken in the middle of the night after fasting for ∼5-6 h. The calcium supplement was usually taken in the afternoon or evening.
BMD was measured at the spine and hip by DXA using a Hologic QDR1000 densitometer (Sheffield), three QDR2000plus densitometers (Sheffield, Manchester, and Newcastle), a Lunar DPX-L (Manchester), and a Lunar Expert 1129 (Liverpool). The European Spine Phantom was circulated to all centers to allow computation of standardized BMD (sBMD) and allow comparison of between-center differences at study entry. Measurements for each individual were made on the same densitometer throughout the study, and the change from baseline was expressed as a percentage without converting to sBMD. Forearm BMD was measured at the Sheffield center using single-energy X-ray absorptiometry (Osteometer DTX100). BMD was measured at baseline and at 6-monthly intervals thereafter.
At the two largest sites (Universities of Sheffield and Manchester), broadband ultrasound attenuation (BUA) and speed of sound (SOS) were measured by quantitative ultrasound (QUS) of the calcaneus. At both sites, these measurements were made in the nondominant heel using the Cuba Clinical (McCue PLC, Southampton, UK) with measurements captured from baseline to 12 months. Maximum speed of ultrasound was also measured at the midpoint of the nondominant tibia (tibial SOS) using the SoundScan 2000 (Myriad, Rehovot, Israel).(14)
Vertebral fractures were assessed on lateral spine radiographs at baseline and at annual intervals thereafter. Prevalent vertebral fractures at entry were assessed by vertebral morphometry using a semiautomated method, and incident vertebral fractures were defined using the same method (point prevalence, i.e., any incident fracture had to be identified as a prevalent fracture on the follow-up radiographs) together with a required minimum change in vertebral height from the baseline films of at least 15% and of at least 4.6 mm in absolute height.(13) Nonvertebral fractures were captured as adverse events. Comparison between the study groups excluded fractures caused by high trauma or fractures at sites deemed to be non-osteoporotic (face, hands, feet, and ankles).(15)
At baseline and 6, 12, 24, and 36 months, blood and urine were obtained after an overnight fast. In a subset of patients (N = 53 and N = 58 in the clodronate and placebo groups, respectively), bone turnover was estimated by measuring total and bone-specific alkaline phosphatase (ALP; HiCORE alkaline phosphatase and BONE-ALP, respectively; Roche Diagnostics) as formation markers and serum c-telopeptide of type I collagen as a marker for resorption (CTX; Serum Crosslaps; Osteometer BioTech A/S). The ALP measurements were done on a Hitachi 717, with all analyses run as single determinations.
Measurements of the safety of treatment included serum calcium (adjusted for albumin), serum phosphate, and indices of hematological, hepatic, and kidney function at entry and at regular intervals thereafter. Information about adverse events was also collected at regular intervals throughout the study.
Standardized BMD was computed for all scanners, except the Lunar Expert, using the formulae published by Pearson et al.(16) sBMD for the Expert was computed using a cubic nonlinear regression model. All results are expressed as the mean ± SEM unless stated otherwise.
Comparison of baseline characteristics of the two study strata was carried out using the χ2 test and ANOVA where applicable. Analysis of the effect of clodronate on efficacy variables was carried out on an intention-to-treat basis. For the analyses of differences in BMD and safety indices, the ANOVA of repeated measures was used. If the assumptions of parametric testing were not met, transformations were used. If none of them proved to be appropriate, nonparametric analyses were applied. χ2 tests were used to determine the statistical differences in the effects of clodronate and placebo on the incidence of vertebral fracture. The preplanned primary outcome for this analysis was the effect of treatment on the proportion of patients sustaining incident vertebral fractures in the two strata combined. The secondary analyses comprised the effect of treatment on BMD in the two strata combined. Results for the individual strata are also reported. In these analyses, significance was considered at p < 0.05.
The relationship between baseline characteristics, the future incidence of vertebral fractures, and the ability of clodronate to reduce fracture risks in various subgroups of patients was examined using simple χ2 tests and logistic regression analyses, including interaction terms for treatment and the baseline characteristic under examination.
A total of 593 women with osteoporosis was recruited to the trial, comprised of 483 women with postmenopausal osteoporosis (stratum I) and 110 women with secondary osteoporosis (stratum II; Fig. 1). The characteristics of patients at entry to the study are shown in Table 1. There were no statistically significant differences between the women with postmenopausal osteoporosis or secondary osteoporosis. The causes of secondary osteoporosis in women included corticosteroid use (n = 73), previous thyrotoxicosis (n = 7), previous nonrecurrent malignancy (breast cancer, n = 11; others, n = 6), hyperparathyroidism (n = 2), and miscellaneous causes (n = 11).
Table Table 1.. Characteristics of the Patients in the Two Strata at Entry to the Study
Similar numbers of women withdrew before the end of the 3-year study in the clodronate and placebo groups (100 [34.2%] and 99 [32.9%], respectively; Fig. 1). In both groups, discontinuations occurred most frequently in the first year of treatment (44 versus 40, respectively), but spine radiographs were available for analysis at 12 months or more in 520 women (88%). Of the women completing 1 and 3 years, repeat spine radiographs were available for analysis of incident vertebral fractures in >99% of subjects remaining in the study at each time point. Withdrawals from the study because of adverse events were slightly, but not significantly, higher in the clodronate treatment group (66 versus 54 women, p = 0.19). In general, treatment with clodronate was well tolerated, with reported side effects confined to the lower gastrointestinal tract. Thus, diarrhea was slightly more common in the clodronate-treated patients (26.7%; 95% CI, 21.6-31.8 versus 18.6%; 95% CI, 14.2-23.0). There were no differences in the incidence of upper gastrointestinal adverse events, including symptoms suggestive of esophagitis or esophagitis proven on gastroscopy.
Effect on the incidence of vertebral fracture
Follow-up radiographs for incident vertebral fractures at 1 year were obtained in 520 (100%) of the women remaining in the study for at least 6 months and at 3 years in 418 women (100%) completing at least 30 months of the study.
Vertebral fractures were present in 49% of the women with postmenopausal osteoporosis and in 60% of the women with secondary osteoporosis at entry to the study. The prevalences of vertebral fractures were similar in the clodronate and placebo groups at baseline (57% versus 52%, respectively).
New vertebral fractures occurred in 63 women in the placebo group during the 3-year study, equating to an incidence of 23.3% (Figs. 2 and 3). In contrast, new vertebral fractures were only observed in 33 patients receiving clodronate, an incidence of 12.7% (relative risk [RR], 0.54; 95% CI 0.37-0.80; p = 0.001; Figs. 2 and 3). The effect was observed in both strata (I: RR, 0.63; 95% CI, 0.40-0.98; p = 0.038; II: RR, 0.35; 95% CI, 0.16-0.76; p = 0.001), with no significant interaction between the treatment and stratum. Treatment was associated with a reduction in both the incidence of single and multiple vertebral fractures. Thus, 8% and 15% of women sustained single vertebral fractures in the clodronate and placebo groups, respectively (RR, 0.56; 95% CI, 0.32-0.96), whereas 3.9% and 7.8%, respectively, had two or more incident fractures (RR, 0.52; 95% CI, 0.25-1.09).
As expected, the incidence of vertebral fracture was higher in those patients with one or more prevalent vertebral fractures at baseline; for example, in the placebo group, the incidence in women with prior vertebral fracture was 38.3% compared with only 8.6% in those without fractures (p < 0.0001). Although the numbers with new fractures were few, the efficacy of clodronate to reduce the incidence of fracture was observed in those without prevalent vertebral fracture (2.3% versus 8.6% for 3 versus 12 patients, respectively; RR, 0.27; 95% CI, 0.08-0.93) and those with prevalent vertebral fracture (22.5% versus 38.3% for 29 versus 48 patients, respectively; RR, 0.59; 95% CI, 0.40-0.87).
The incidence of vertebral fractures was significantly lower by the end of the first year in those women receiving clodronate and remained lower during each year of treatment (Fig. 3). The reduced incidence of vertebral fractures was associated with a significant reduction in standing height loss (0.62 versus 0.97 cm at 3 years in the clodronate and placebo patients, respectively; p = 0.02; Fig. 3).
Effect on nonvertebral fractures
A total of 35 patients sustained osteoporosis-associated nonvertebral fractures during the 3 years of the study. Fractures were documented in 21 women in the placebo group compared with 14 receiving clodronate. The sites of fracture were the wrist/distal forearm (9 and 8 in the placebo and clodronate arms, respectively), the upper arm (5 and 3), the hip (6 and 1), and the lower limb (1 and 2).
Effect on BMD, quantitative ultrasound, and biochemical markers of bone turnover
Mean spine BMD did not differ significantly between the two treatment groups at baseline in the two study strata (Table 1). Hip BMD was lower in the women randomized to receive placebo in stratum II, despite similar spine BMDs.
In all patients, clodronate was associated with a significant increase in mean spine BMD over 3 years (percent change from baseline, 4.35 ± 6.34% versus 0.64 ± 6.02% in the placebo group, p < 0.0001; Fig. 4A). At the hip, clodronate maintained total BMD, whereas a significant decrease was observed in the placebo group (percent change from baseline 0.70 ± 5.67% versus −3.03 ± 6.32% in the placebo group, p < 0.0001; Fig. 4B). There was no significant difference between the treatment effects in the women with postmenopausal or secondary osteoporosis. Over the 3 years, distal forearm BMD showed no significant change from baseline in those women receiving clodronate (+0.58 ± 4.66%), whereas a significant decrease occurred in the placebo group (−2.31 ± 6.67%, p = 0.001; Fig. 4C). Mean changes in heel BUA, but not heel SOS, were also significantly different between the clodronate and placebo groups (p < 0.05). At 12 months, heel BUA had increased by 3.6 ± 12.8% in the clodronate group (n = 59), whereas the mean change in the placebo group was −2.7 ± 14.6% (n = 59). Tibial SOS did not change significantly in either treatment group.
At baseline, the median (interquartile range) of bone-specific ALP (28.5, 20.0-37.0 versus 29.0, 18.0-37.0 U/liter) and serum CTX (1500, 1216-2752 versus 1910, 1265-2513 pM) were similar in those women randomized to receive clodronate and placebo, respectively. Treatment with clodronate was associated with a significant 17% median decrease in serum bone-specific ALP within the first 6 months, whereas values remained unchanged in the placebo group (Table 2). Beyond the first year, both groups showed a tendency for increasing levels of bone-specific ALP, but the median changes from baseline remained significantly lower in the clodronate group (Table 2). A similar pattern was observed using serum CTX (Table 2), where a 46% median decrease in values was observed during the first 6 months of clodronate treatment, whereas values remained stable in the placebo group. Values of serum CTX showed an upward trend from 6 to 12 months in both groups, but the gradient of increase was much steeper in the placebo group; therefore, treatment with clodronate was associated with significantly lower values at all time points during the 3 years of the study (Table 2).
Table Table 2.. Median Percentage Changes From Baseline Values in Biochemical Markers of Bone Turnover
Interaction between baseline variables, fracture incidence, and clodronate efficacy
Women sustaining one or more incident vertebral fractures during the 3-year study were significantly older than those women remaining fracture-free (Table 3). They also had a significantly higher prevalence of vertebral fractures at entry, reflected in a significantly lower standing height. Other skeletal indices were also significantly different between the fracture and nonfracture groups, namely a more marked reduction in spine and hip sBMD at baseline (Table 3). Heel BUA and SOS tended to be nonsignificantly higher at baseline in those women who remained free of new vertebral fractures. Mean values of serum bone-specific ALP and CTX were similar in the two groups. Differences in nonskeletal risk factors included a significantly higher prevalence of corticosteroid use in the fracture group and a nonsignificant increase in the reporting of current or ever smoking.
Table Table 3.. Comparison of Patients Sustaining One or More Incident Vertebral Fractures to Those Remaining Free From Vertebral Fracture Over 3 Years
The interactions between these baseline variables, the incidence of fracture, and the efficacy of clodronate to reduce or prevent fractures are shown in Table 4. None of the baseline variables showed a significant interaction with the effect of clodronate to reduce the incidence of vertebral fractures. For example, in those women reporting current or previous smoking, the point estimate of the odds ratio for fracture incidence during clodronate treatment was 0.47 compared with 0.46 in those who had never smoked (Table 4). Likewise, the presence of lower values of BUA was not associated with any significant reduction in the efficacy of clodronate. It is of interest to note that, when women were classified by total hip BMD at study entry, the effect of clodronate to reduce fracture risk was observed in women with osteopenia (sBMD lying between 640 and 820 mg/cm2—corresponding to T-scores of −2.5 and −1, respectively), as well as those with osteoporosis (Table 4). The event rates in those women with normal BMD were too small to exclude efficacy within this group.
Table Table 4.. Women With Incident Vertebral Fractures by Treatment Groups and Risk Subgroups (OR for Vertebral Fractures in Clodronate Group Compared With the Placebo Group by Risk Subgroups)
Effect of clodronate on biochemical and hematological indices of safety
Treatment with clodronate was well tolerated. There were no clinically significant effects of clodronate on hematological or biochemical indices of safety. Mean values of leukocytes and platelets tended to be lower during treatment in women receiving clodronate, but the changes were small, the mean values remained well within the normal range, and no clinically significant abnormalities occurred. Mean serum creatinine values lay within the normal range in both groups at entry and tended to show minor but statistically significant increases in both groups over time. There was no significant difference between the clodronate and placebo groups. Clodronate induced a small but statistically significant decrease in mean serum calcium values (−0.02 to −0.03 mM) within the normal reference range, whereas values remained stable in the placebo group. Serum phosphate also showed a small but clinically insignificant decrease in the clodronate group. Finally, clodronate was associated with small, statistically significant but clinically unimportant, increases in serum transaminases.
This is the first adequately powered study to investigate the effect of clodronate on vertebral fracture incidence in women with osteoporosis. These results over a period of 3 years confirm our previous observations that clodronate, when given as a dose of 800 mg daily, is effective in reducing the risk of vertebral fracture in patients with prevalent vertebral fracture or densitometrically defined osteoporosis.(13) Bearing in mind that no head-to-head comparisons have been undertaken, clodronate seems similar to alendronate and risedronate, both of which have had proven reduction of vertebral fracture in clinical trials of 3-4 years of duration.(17–20) In contrast to these studies and the vertebral fracture study of raloxifene,(21) our patients were mainly recruited from those presenting to hospital-based osteoporosis out-patient departments rather than the general population and are therefore typical of patients presenting for management of osteoporosis. Furthermore, by concentrating on patients who have presented with problems associated with osteoporosis, we have achieved statistically significant fracture reduction using a much smaller patient population than was necessary for the previous studies.
As we have previously demonstrated, the reduction in vertebral fracture risk was evident within the first year of treatment.(13) The results add further evidence to the observation that all antiresorptive therapies have a similar rapid effect to reduce the incidence of spinal fractures.(20,22) These results show that this reduced fracture risk continued through 3 years of treatment with clodronate. For example, the RR of new vertebral fracture after clodronate treatment was 0.54 at 3 years, which is identical to that observed after only 1 year of clodronate therapy. This is similar to results reported for raloxifene(21,23) but contrasts with observations in studies of risedronate, which suggest that the major effect of treatment may occur within the first year.(19) The reduction in vertebral fracture incidence is mirrored in the significant reduction in height loss, an important clinical consequence of vertebral fracture that contributes to morbidity.(24,25) The ability of clodronate to reduce vertebral fracture risk and height loss has also been observed in patients with multiple myeloma,(4) and similar observations have been reported in male osteoporosis with alendronate and after renal transplantation with ibandronate.(26,27)
The ability of clodronate to reduce vertebral fracture risk is independent of several baseline factors known to influence fracture risk. For example, treatment is effective in the presence or absence of baseline vertebral fractures. Despite an apparently greater effect to reduce fracture risk in women without fracture at entry, the absolute risk reduction is much larger in those with prevalent fractures (15.8% versus 6.3%), resulting in a lower number needed-to-treat (NNTs of 6 and 16, respectively, over 3 years to prevent one woman having a vertebral fracture). Age, corticosteroid use, smoking habits, weight, and QUS parameters had no detectable impact (Table 4) on the beneficial effect of clodronate. In particular, low values of QUS in patients with low BMD and/or prevalent vertebral fractures, possibly reflecting lower bone quality, do not seem to lessen the response to antiresorptive therapy. It is therefore unlikely that any of these will be helpful in identifying patients who will benefit especially from clodronate therapy. On the other hand, these factors are all markers of the risk of future osteoporotic fracture and therefore may be useful in the identification of patients who are at an increased risk of fracture and may benefit from therapy with clodronate to reduce that risk. The role of QUS and other peripheral measures of skeletal strength as the primary means of identifying patients for antiresorptive therapies needs to be established in appropriate clinical trials.
Recent experience with alendronate and risedronate has suggested that fracture reduction may only be seen in patients with low BMD, confirmed by DXA, before treatment.(17,28) Our study suggests that antifracture efficacy is independent of the underlying BMD, a finding that is consistent with other recent studies.(29,30) Several studies suggest that changes in bone turnover, rather than changes in BMD, determine responsiveness to antiresorptive therapies, at least at the spine,(31,32) if not at nonvertebral sites.(33,34) The effect of clodronate on spine BMD in our study is very comparable with that observed in the recent placebo-controlled dose comparison study of Valimaki et al.,(12) with between-group differences of ∼4% at the end of 3 years. These changes in spine BMD are intermediate to those observed in the studies of raloxifene(21) and alendronate.(17,18) Likewise, the effect of clodronate on serum bone specific ALP seems intermediate to that of raloxifene(35) and risedronate(36) but is somewhat less than that of alendronate.(36) Nonetheless, the effect to reduce vertebral fracture is comparable across the range of agents. Because of our study design, some patients were enrolled with a BMD above the osteoporosis threshold if vertebral fractures were present, and there certainly doesn't seem to be any significant difference in the efficacy of clodronate across the ranges of total hip BMD. Biochemical markers of bone turnover were not measured at entry in all subjects, but we have previously demonstrated that baseline serum ALP was significantly higher in those with prevalent vertebral fractures.(37) It is not possible to be sure whether the increased bone turnover markers preceded the fractures or simply reflected the response to recent fractures, but the efficacy of clodronate suggests that the former may be true. No study has yet been reported that has taken a random population sample, regardless of BMD, to formally determine whether antiresorptive therapies are able to reduce fracture incidence in patients with BMD values lying above or below the WHO threshold. A large community-based study in elderly women living in the United Kingdom (the MRC/Leiras-funded HIPS study) has been designed to address this and other issues and is expected to report in the near future.
Clodronate treatment was well tolerated. There was no suggestion of upper gastrointestinal intolerance, which has been noted with other bisphosphonates, particularly those that possess a nitrogen-containing side chain, which is not present in clodronate.(38) There was a slight increase in diarrhea in patients receiving clodronate, but this was seldom of any clinical significance and probably reflects a purgative effect of the phosphate delivered to the colon by the clodronate molecule. Similar disturbances have been noted in studies of clodronate in primary breast cancer(6) and have been reported with other non-nitrogen-containing bisphosphonates.
We conclude that clodronate 800 mg daily is an effective therapy to increase bone density and reduce the risk of vertebral fracture in women with postmenopausal or secondary osteoporosis. Clodronate is a potential alternative to other established therapies for postmenopausal osteoporosis.
We thank Linda Reaney for technical skills in performing the vertebral height measurements and all members of the scanning and nursing departments who provided much needed assistance in conducting the study. This study was funded by Leiras Oy, Finland.