Ten RCTs (1,497 patients) were included in the review. The sample size ranged from 13 to 643.
Nine trials were placebo-controlled, of which seven described double-blinding. Two trials described randomisation methods. Four trials estimated sample size and described study power. Two trials performed analysis according to intention-to-treat. Patient follow-up and details of those who had completed treatment as per protocol or had withdrawn were rarely reported.
Four of five clodronate trials did not detect significant differences in pain outcomes between clodronate and placebo. The fifth had some methodological problems (including having just 2 weeks’ follow-up of 13 patients). No statistically significant differences in mean pain scores were detected between treatment arms at 9 and 27 weeks in a pooled analysis of two pamidronate trials. Statistically significantly lower increases in pain scores were observed when comparing zoledronic acid with placebo in the only trial to investigate this.
The trial investigating zoledronic acid found a statistically significant reduction in the number of patients having at least one SRE after 15 months and 24 months of treatment at a dose of 4 mg. The pamidronate trials found no difference in rates of SRE. Other trials did not report this outcome.
None of three trials investigating survival detected a statistically significant difference between treatment groups (two trials of clodronate and one of zoledronic acid).
Two trials investigated quality of life. One of zoledronic acid found no statistically significant difference between groups and one of clodronate found improvements compared with placebo only in the pain domain.
Overall, bisphosphonates were well tolerated with the majority of trials finding similar proportions of mild toxicity between bisphosphonate and placebo. Greater proportions of adverse events in comparison with placebo were observed in the zoledronic acid trial.