Could Strontium Ranelate Have a Synergistic Role in the Treatment of Osteoporosis?

In the current issue of the Journal, Recker et al.⁽¹⁾ present results from a randomized active comparator study in osteoporotic postmenopausal women that provide compelling evidence that treatment with strontium ranelate (SrR) has only a modest effect on bone remodeling as assessed by biochemical markers of bone turnover and bone histomorphometry. The authors report that changes in biochemical markers during 6-mo treatment with SrR failed to confirm any bone-forming activity. They conclude that the modest effect of SrR on bone remodeling and cell activity may indicate that its effects on fracture reduction are mediated by a different mechanism than that observed with anabolic or the more potent antiresorptive agents. In this study, we discuss whether, if SrR works differently to other treatments, its use in combination with other osteoporosis therapies might have a synergistic effect, with greater benefit to the patient than the use of a single treatment alone.

Over the past two decades, the range of therapeutic options for osteoporosis has increased dramatically.^(2–11) Most of the available drugs act through their effects on bone remodeling. Antiresorptive drugs such as the bisphosphonates and selective estrogen receptor modulators reduce osteoclast development and activity, resulting in lower bone turnover. The consequence of this is an increase in BMD caused by a reduction in the formation of new remodeling units, in-filling of the existing remodeling space, and an increase in mineralization.^(12,13) The reduction in fracture risk is greater than can be explained by the relatively modest changes in BMD,⁽¹⁴⁾ most probably reflecting suppression of high bone turnover, which seems to be an independent risk factor for fracture.⁽¹⁵⁾ Anabolic skeletal agents such as the PTH peptides increase bone formation and produce much larger increases in BMD, at least in the spine.⁽⁷⁾ The mechanisms by which PTH peptides increase bone mass are incompletely understood, but increased bone formation at the tissue and cellular level results in a combination of de novo bone formation by modeling and increased formation in basic multicellular units during remodeling.⁽¹⁶⁾

In contrast to these interventions, there is evidence that the antifracture effects of SrR are mediated, at least in part, through mechanisms that are independent of bone remodeling.⁽⁹⁾ Initially, it was proposed that this compound acted by uncoupling the remodeling process by simultaneously stimulating bone formation and suppressing bone resorption.^(17–20) However, this contention, mainly based on animal experiments, has recently been challenged.⁽²¹⁾ In normal rats, mice, and monkeys, a dose-dependent decrease in bone resorption and increase in bone formation, resulting in increased bone mass and mechanical strength, has been reported.⁽¹⁹⁾ In ovariectomized or immobilized rats SrR was shown to prevent bone loss, decreasing bone resorption but not bone formation, suggesting that it may act as an uncoupling agent.⁽²⁰⁾ In both models of bone loss, a dose-dependent increase in plasma alkaline phosphatase activity was shown, indicating stimulatory effects on bone formation.^(17,19) However, in another study at similar doses of SrR, no stimulatory effects were observed in ovariectomized rats, nor could improvements in bone biomechanical properties be shown.⁽²¹⁾ In vitro studies have provided some evidence for inhibitory effects of SrR on osteoclasts and stimulatory effects on osteoblasts.⁽²⁰⁾

Even before publication of the study by Recker et al.,⁽¹⁾ data from postmenopausal women treated with SrR failed to provide convincing evidence that the effects of the drug on fracture risk are mediated by changes in bone remodeling. Measurements of biochemical markers in treated women have yielded inconsistent results. In a 2-yr randomized controlled trial of SrR (125 mg, 500 mg, or 1 g/d) in early postmenopausal women, serum osteocalcin (OC) levels increased in both the placebo and SrR-treated groups, and no significant treatment effect could be shown, nor was there any significant effect of SrR on urinary cross-linked C-terminal telopeptide, a marker of bone resorption.⁽²²⁾ Although statistically significant increases in serum bone-specific alkaline phosphatase (bALP) levels were reported, these were not confirmed by later studies.^(1,9,23,24) Some of the results of this study may have been influenced by the fact that the doses used were lower than the approved dose of 2 g/d.

Similar findings were reported in a phase II study in postmenopausal women with established osteoporosis; significant treatment effects of SrR (2 g/d) on serum bALP were seen, but no consistent or statistically significant effects on serum OC or procollagen were seen for any of the doses administered (500 mg, 1 g, or 2 g/d).⁽²³⁾ However, urinary excretion of type 1 collagen cross-linked N-telopeptide pyridinoline (NTX) showed a small but significant decrease relative to placebo of ∼20% in the two highest dose groups over the first 6 mo.

In the phase III SOTI (Spinal Osteoporosis Therapeutic Intervention) study, treatment with 2 g/d of SrR in postmenopausal women with established osteoporosis was associated with only a small, albeit statistically significant, increase in serum bALP levels compared with placebo (∼8%) at 3 mo and thereafter for the 3-yr duration of the study.⁽⁹⁾ Although serum C-telopeptide cross-link levels were significantly lower in the SrR-treated group compared with placebo, the magnitude of this difference was relatively small (∼12%), and after 12 mo, there were increases above baseline in this resorption marker in both the treatment and placebo groups.

In the new study reported by Recker et al.,⁽¹⁾ postmenopausal women with osteoporosis were randomized to receive 6-mo treatment with either human recombinant PTH (teriparatide 20 μg/d administered subcutaneously) or oral SrR (2 g/d). Treatment with teriparatide induced large and highly significant increases from baseline in bone formation and resorption markers, reaching statistical significance for serum amino-terminal propeptide of type 1 collagen (PINP) after 1 mo (+57%, p < 0.001). In contrast, the corresponding changes in the SrR-treated group were modest with small but statistically significant reductions in PINP at 3 (−14%, p < 0.005) and 6 mo (−19%, p < 0.001) and in the bone formation marker serum β-C-terminal telopeptide of type-1 collagen (β-CTX) at 1 and 3 mo (−11% for both, p < 0.05). No changes were found in bALP. The authors interpreted the overall changes from baseline in the SrR-treated group as consistent with a weak antiremodeling effect.

Previous histomorphometric data in humans have also failed to provide convincing evidence for changes in bone remodeling commensurate with the effects of SrR on bone strength.⁽²⁵⁾ No significant differences in indices of bone turnover were reported between biopsies from women receiving SrR and placebo, arguing against a significant antiresorptive effect. The absence of any significant change in the degree of mineralization of bone treated with SrR also argues against substantial effects on bone remodeling. Higher mineral apposition rate and osteoblastic surface in cancellous bone of women treated with SrR would be consistent with increased bone formation, but these changes were small, and no significant increase in cancellous bone volume or trabecular thickness was shown. In the study of Recker et al., the lack of paired bone biopsies and nonevaluable samples in some subjects reduced the power of the study to show statistically significant differences in histomorphometry findings between the teriparatide and SrR groups.⁽¹⁾ However, the majority of the bone formation parameters in the teriparatide group were numerically greater than in the SrR group, with some differences approaching statistical significance. The authors conclude that their data provide no evidence that SrR has any bone anabolic activity.

If the antifracture effects of SrR cannot be explained on the basis of changes in bone remodeling, what is the mechanism of action? Treatment with SrR leads to large increases in BMD, comparable in the spine to that seen with PTH peptide therapy, and significantly larger at the hip.⁽⁹⁾ During treatment, some strontium is incorporated into hydroxyapatite, where it can substitute for up to 1 in 10 calcium atoms.⁽²⁶⁾ Hence, much of the BMD increase is likely to be a technical artefact because strontium has a higher atomic number than calcium, with greater attenuation of the X-rays used for DXA measurements.^(27,28) Studies of bone biopsy specimens from the iliac crest show that, after 3 yr of treatment, ∼1.6% of calcium atoms in bone are replaced by strontium,⁽²⁹⁾ and this probably accounts for between 75% and 100% of the measured BMD increase.⁽³⁰⁾ The greater breaking strength of bone treated with SrR may therefore be caused by the incorporation of strontium into the hydroxyapatite crystal, so that the mechanism of action is largely caused by a physical effect associated with the increase in bone hardness measured by nanoindentation.⁽³¹⁾ There is some evidence that the reduction in fracture risk in SrR-treated patients may be greater in those who experience a larger BMD increase⁽³²⁾ and, because BMD change is principally a measure of bone strontium content,⁽³⁰⁾ this could be construed as evidence supporting the hypothesis of a physical effect of strontium on bone strength.

If the treatment effect of SrR is explained by bone strontium content, the long-term retention of strontium in the skeleton for many years after the withdrawal of treatment⁽²⁸⁾ could be beneficial. Results of the 5-yr SOTI extension study in subjects who stopped active treatment at 48 mo showed that, on average, 20% of the gain in spine and hip BMD during the previous 4 yr was lost in the first 12 mo after stopping therapy,^(33,34) a result consistent with the predictions of theoretical modeling based on the International Commission on Radiological Protection whole body strontium retention function.^(28,34) The same model predicts that bone strontium content will decrease to 50% of the level achieved at the end of therapy by 5 yr after stopping treatment and to 33% at 10 yr.⁽²⁸⁾ It is therefore possible that patients previously treated with SrR might continue to derive at least some antifracture benefit for a period of 5 or 10 yr after discontinuing therapy.

Physicians treating osteoporosis therefore have several therapeutic approaches that differ in their mechanisms of action. Under the optimal conditions of clinical trials, there is generally an ∼50% reduction in vertebral fractures and, for some interventions, a reduction of ∼20% in nonvertebral fractures.^(2–11) In clinical practice, the results are likely to be worse, and a large number of osteoporotic fractures continue to occur. Combining drugs with different mechanisms of action could decrease or increase efficacy; for example, in the case of combined anabolic and antiresorptive therapy, efficacy is decreased, at least in terms of BMD changes,⁽³⁵⁾ and there is also evidence that the anabolic effects of PTH peptides may be blunted by prior antiresorptive therapy.^(36,37) Conversely, the use of SrR after anabolic therapy might produce additional biomechanical benefits, the physico-chemical effects on bone mineral providing the equivalent of a “medical vertebroplasty.” Similarly, antiresorptive therapy after treatment with PTH peptides may not only maintain BMD⁽³⁸⁾ but might also increase bone strength as a result of increased mineralization.⁽³⁹⁾ Finally, combined or sequential use of SrR and bisphosphonates provides interesting potential approaches, given the lack of convincing evidence for significant effects on bone remodeling of the former. Thus, combined or sequential therapy with SrR should not reduce the beneficial effects of bisphosphonates on bone turnover, and the combined effects of the two drugs on bone strength might be synergistic. These options have yet to be explored but could provide novel approaches to the treatment of osteoporosis and further clarify the mechanism of action of SrR on bone.