The randomized phase III trial of zoledronic acid involved patients with bone lesions associated with breast cancer or multiple myeloma, and was a noninferiority trial directly comparing zoledronic acid with pamidronate, the standard of care at the time. Like the pamidronate trials previously discussed, the zoledronic acid trial assessed efficacy using the composite SRE end-point, and evaluation of the primary end-point (percentage of patients with at least one SRE) excluded hypercalcemia from the definition of SREs. In addition, this trial included a planned multiple event analysis using the Andersen-Gill method,29 which provides a comprehensive assessment of skeletal morbidity based on both the incidence and timing of all SREs. Bone pain was assessed by measuring mean change from baseline in brief pain inventory (BPI) scores.

Results from this study demonstrated that zoledronic acid (4mg) was at least as effective as pamidronate (90mg) in the overall patient population and was significantly more effective at reducing the risk of developing an SRE (by Andersen-Gill analysis) compared with pamidronate in the overall patient population.9,30 Moreover, zoledronic acid demonstrated superiority to pamidronate in the subset of patients with breast cancer.A total of 766 breast cancer patients with bone metastases were randomized to zoledronic acid (4mg via 15-minute infusion) or pamidronate (90mg via two-hour infusion) every three to four weeks for up to 24 months. At the final analysis at 25 months, the percentage of patients who developed at least one SRE was similar between treatment groups (46% for zoledronic acid versus 49% for pamidronate), which is consistent with the results of the placebocontrolled pamidronate trials previously discussed (see Figure 2).3,6,9 However, zoledronic acid consistently reduced the percentage of patients with each type of SRE,16 and the Andersen-Gill multiple event analysis showed that treatment with zoledronic acid significantly reduced the overall risk of experiencing an SRE by an additional 20% compared with pamidronate (hazard ratio=0.799; p=0.025).9 Both zoledronic acid and pamidronate also decreased pain scores in this trial. Among patients with pain scores >0 at baseline, mean composite BPI pain scores decreased from baseline at every time-point up to month 13 (the last time-point assessed) regardless of treatment.30 Moreover, at 25 months, significantly fewer patients treated with zoledronic acid required radiotherapy to bone, which can be considered a surrogate for bone pain, compared with pamidronate (19% versus 27% for pamidronate; p=0.011).16 Importantly, zoledronic acid (4mg) was well tolerated; flu-like symptoms were more common in patients treated with zoledronic acid compared with placebo,but these were mild and occurred primarily after the first infusion.

Recently, a randomized, placebo-controlled study was conducted to investigate the efficacy of zoledronic acid in 228 Japanese women with bone metastases from breast cancer.10 In this trial, patients were treated with zoledronic acid (4mg every four weeks) or placebo for one year. The primary end-point in this trial was the ratio of the SRE rate (defined as the total number of SREs divided by time on study) for patients treated with zoledronic acid versus placebo. Similar to the other trial, this primary efficacy analysis did not include hypercalcemia in the definition of SREs. Results demonstrated that zoledronic acid significantly reduced the SRE rate by 43% compared with placebo (p=0.016). Zoledronic acid also significantly reduced the percentage of patients who developed an SRE (31% versus 52% for placebo; p=0.001), delayed the onset of skeletal complications (p=0.004), and reduced the overall risk of developing an SRE by 44% (hazard ratio =0.56; p=0.009) compared with placebo. Moreover, zoledronic acid provided durable reductions of mean BPI pain scores from baseline and compared with the placebo group throughout the study. Patients treated with zoledronic acid experienced a statistically significant decrease from baseline in their mean pain score at every time-point beyond week two, including the 12-month time-point (p=0.0004).