What Else Do We Need? A Commentary on Zoledronate Effects on Cancer and Cardiac Events

The devastating consequences of skeletal fragility are well known, as are the unequivocal benefits of prescribing osteoporosis medications to patients who are at increased risk of fracture. Nonetheless, those of us who focus on treating osteoporosis are forced to spend much of our time on the defensive because we are regularly challenged to justify our treatment decisions even when the clinical scenario is straightforward. These challenges come from many directions, including patients and referring physicians, and tend to focus on concerns over rare side effects, long-term efficacy, and the potential to reduce patient mortality. And although these concerns are legitimate in certain circumstances, in appropriately selected patients in whom the necessity of continuing treatment is continually reassessed, they are not only misplaced but also detrimental to public health. In this setting, the article published in the current edition of the Journal of Bone and Mineral Research (JBMR), entitled Effects of Zoledronic Acid on Cancer, Cardiac Events and Mortality in Osteopenic Women,1 has the potential to provide us with additional tools as we work to close the osteoporosis treatment gap and reverse current negative trends.

The potential of the nitrogen-containing bisphosphonates (by far the most commonly used class of osteoporosis medications) to reduce mortality has been an area of significant interest since their introduction in the 1990s. Specifically, numerous observational studies have assessed the capacity of bisphosphonates (primarily alendronate) to reduce mortality and have generally reported a significant relationship between bisphosphonate use and lower death rates.2-6 Still, appropriate skepticism has persisted because of the concern that these types of pharmacoepidemiologic investigations are susceptible to confounding.7 Importantly, however, meta-analyses of randomized controlled trials have also supported an overall mortality benefit as well as a specific reduction in cardiovascular risk.8, 9 Based on the US Food and Drug Administration's requirements for drug approval, however, most individual studies of osteoporosis drugs have not been powered to assess survival but rather were specifically designed to efficiently demonstrate a reduction in fracture incidence. Nonetheless, to date there have been two randomized double-blind placebo-controlled trials that have suggested a mortality benefit with bisphosphonate therapy, both of which involved intravenous zoledronic acid.10, 11 In the more recent of these two studies, Reid and colleagues11 conducted a 6-year, double-blind trial involving 2000 osteopenic women aged 65 years or older randomly assigned to receive four infusions of either 5 mg zoledronic acid or saline every 18 months and reported not only a decreased risk of vertebral and nonvertebral fractures, but also a trend toward an overall decrease in death (odds ratio of 0.65; 95% CI, 0.40 to 1.05).

In this issue of the JBMR, Reid and colleagues1 now provide a critical in-depth analysis of the adverse events experienced by patients in their previous study.11 Specifically, study subjects recorded events in a questionnaire every 3 months and hospital admissions were reviewed by a study physician. Certain adverse events were prespecified for analysis (without adjustment for multiple comparisons), including death, sudden death, myocardial infarction, coronary revascularization procedures, stroke, transient ischemic attacks, cancer, and a combination of all vascular events. The findings were striking. Specifically, the incidence of myocardial infarction in the placebo group (43 events in 39 women) was higher than in the zoledronic acid group (25 events in 24 women; hazard ratio 0.60 [95% CI, 0.36 to 1.00; and rate ratio 0.58 [95% CI, 0.35 to 0.94]). Similarly, for the composite endpoint of all vascular events, 69 women experienced at least one vascular event in the placebo group compared to 53 women in the zoledronic acid group (hazard ratio 0.76 [95% CI, 0.53 to 1.08] and rate ratio 0.72 [95% CI, 0.53 to 0.98]). Stroke incidence did not differ significantly between groups but fatal stroke did (seven in the placebo group versus one in the zoledronic acid group, p = 0.04). Overall cancer incidence also differed between groups with 127 women diagnosed in the placebo group compared to 87 women in the zoledronic acid group (rate ratio 0.68 [95% CI, 0.52 to 0.89]) as did breast cancer incidence (rate ratio 0.59 [95% CI, 0.35 to 0.98]). Notably, for all of these endpoints, the pattern of events suggested that the protective effects begin within the first year of treatment and then were progressive during the remaining 6-year study period. Although Reid and colleagues1 are careful to point out the limitations of their analysis, including the fact that their study was not specifically powered to show differences in these safety endpoints and they did not correct for multiple comparisons, they also point out that their findings are generally consistent with prior observation studies and meta-analyses in terms of overall death, cancer incidence (and breast cancer specifically), and vascular events. Moreover, as the authors discuss, physiologically plausible mechanistic explanations for these protective effects continue to be actively investigated.

So now, considering the studies by Reid and colleagues1, 11 and all the work that has come before it, the question remains: What additional information do we need to forcefully educate our patients and colleagues, reverse current trends, and make osteoporosis treatment a greater priority in our healthcare system? To answer this question, we must first recognize a few truths about clinical research: (i) there will often be clinical trial findings that conflict in some details because of variations in study design or differing patient populations; (ii) randomized placebo-controlled trials will rarely address all patient subgroups of interest; and (iii) after the introduction of a new medication or class of medication, new data will continue to emerge that may alter our calculations in terms of risks and benefits. But having granted these points, the answer to the question is still clear—we have what we need now. Specifically, we know that treating osteoporosis reduces fractures and saves lives. And the fact that the precise mechanisms responsible for the positive nonskeletal properties of bisphosphonates are not fully defined and we do not know whether bisphosphonate (or even specific bisphosphonates) are unique in their ability to influence survival should not discourage us from advocating for treatment in appropriate patients. Of course, there is more work to do and in the future it may be possible to tailor our therapies in a true “precision medicine” approach that includes choice of agent, duration of treatment, and drug sequence. But in the interim, the fact that there is more work to do cannot discourage us from clearly presenting the positive balance between the benefits and risks of osteoporosis medications. The results presented in Reid and colleagues1 should now be a significant part of that discussion.