Background: Bone is the most common site of metastatic disease associated with breast cancer affecting more than half of women during the course of their disease. Bone metastases are a significant cause of morbidity due to pain, pathological fractures, hypercalcaemia and spinal cord compression, and contribute to mortality. Bisphosphonates, which inhibit osteoclast-mediated bone resorption, are standard care for tumour-associated hypercalcaemia, and have been shown to reduce bone pain, improve quality of life, and to delay skeletal events and reduce their number in patients with multiple myeloma. Several randomized controlled trials have evaluated the role of bisphosphonates in breast cancer.

Objectives: To assess the effect of bisphosphonates on skeletal events, bone pain, quality of life and survival in women with early and advanced breast cancer.

Search strategy: Randomized controlled trials were identified using the specialized register maintained by the Cochrane Breast Cancer Group (the search was applied to the databases Medline, Central/CCTR, Embase, CancerLit, and included handsearches from a number of other relevant sources). See: Cochrane Collaboration Collaborative Review Group in Breast Cancer search strategy.

Selection criteria: Randomized controlled trials evaluating skeletal events in women with metastatic breast cancer and early breast cancer comparing: 1. treatment with a bisphosphonate with the same treatment without a bisphosphonate 2. treatment with one bisphosphonate with treatment with a different bisphosphonate.

Data collection and analysis: Studies were selected by two independent reviewers. Studies fulfilling the eligibility criteria were evaluated for quality, particularly concealment of allocation to randomized groups. Data were extracted from the published papers or abstracts independently by the two primary reviewers for each of the specified endpoints (skeletal events, bone pain, quality of life and survival). Data on skeletal events and survival were presented as numbers of events, risk ratios and ratios of event rates. Meta-analyses were based on the fixed-effects model (Mantel-Haenszel). Subjective qualitative ratings were used to summarize the quality of life and pain data.

Main results: Twenty one randomized studies were included. All studies in advanced breast cancer included women with clinically evident bone metastases (osteolytic and/or mixed osteolytic/osteoblastic) by plain xray and/or radionucleotide bone scans. In nine studies that included 2189 women with advanced breast cancer and existing bone metastases, bisphosphonates reduced the risk of developing a skeletal event by 17% (RR 0.83; 95% confidence interval (CI) 0.78-0.89; P < 0.00001). This effect was more modest, but still highly significant if episodes of hypercalcaemia were excluded (10 studies, 2656 women, RR 0.85; 95% CI 0.79-0.91 P = 0.0001). Overall, intravenous bisphosphonates reduce the risk of developing a skeletal event by 17 % (95% CI 0.78-0.89) compared with oral bisphosphonates, which reduce the risk of developing a skeletal event by 16 % (95% CI 0.76-0.93). Of the currently available bisphosphonates, 4 mg IV zolendronate reduces the risk of developing a skeletal event by 41% (RR 0.59, 95% CI 0.42-0.82), compared with 33 % by 90 mg IV pamdronate (RR 0.77, 95% CI 0.69-0.87), 18 % by 6 mg IV ibandronate (RR 0.82, 95% CI 0.67-1.00), 14 % by 50mg oral ibandronate (RR 0.86, 95% CI 0.73-1.02) and 16 % by 1600 mg oral clodronate (RR 0.84, 95% CI 0.72-0.98). Compared with placebo or no bisphosphonate, with bisphosphonates the skeletal event rate was lower in all of 12 studies in women with clinically evident bone metastases (median reduction of 29%, range 14-48%); statistically significant reductions were reported in 10 trials (four intravenous pamidronate, two oral clodronate, one intravenous ibandronate and two oral ibandronate, a single intravenous zolendronate study). Studies of intravenous zolendronate, pamidronate and oral clodronate in women with advanced breast cancer and clinically evident bone metastases showed significant delays in the median time to a skeletal event. Event-free survival was also reported to be longer in women receiving 6 mg of ibandronate compared with controls. Compared with placebo or no bisphosphonate, with bisphosphonates significant improvements in bone pain were reported in seven studies (90 mg iv pamidronate, 4 mg iv zolendronate, 6 mg iv ibandronate, 1600 mg oral clodronate and 50 mg oral ibandronate). Eight studies tested the effect of bisphosphonates compared with placebo on patient-rated quality of life using a referenced scale. Improvements in global quality of life were reported in only the three studies of iv and oral ibandronate. Treatment with bisphosphonates does not appear to affect survival in women with advanced breast cancer. Intravenous zolendronate (4 mg) appeared to be as effective as pamidronate (90mg) when directly compared in a single randomized double-blind study, based on the risk of developing a skeletal related event, the median time to first skeletal event and skeletal morbidity rate (events per year). Updated re-evaluation of the primary data in the overall population, by multiple event analysis using the method of Anderson-Gill, showed a reduction in the risk of developing any skeletal complication (including hypercalcamia) of 20 % (zolendronate 4 mg compared with pamidronate 90 mg, RR = 0.80, 95% CI 0.66 - 0.97, p = 0.025), suggesting a possible advantage of zolendronate 4 mg compared with pamidronate 90 mg. In the three studies of bisphosphonates in 320 women with advanced breast cancer without clinically evident bone metastases, there was no significant reduction in the incidence of skeletal events (RR 0.99; 95% CI 0.67-1.47; P = 0.97). In the three studies of oral clodronate that included 1653 women with early breast cancer, there was no statistically significant evidence of reduction in the risk of developing skeletal metastases (RR 0.82; 95% CI 0.66-1.01; P = 0.07), or of visceral metastases (RR 0.95; 95% CI 0.80-1.12, p = 0.53). However there was evidence of improved survival (RR 0.82; 95% CI 0.69-0.97, p = 0.02). However there was statistically significant heterogeneity among these studies and a random effects meta-analysis emphasizes the uncertainty of this finding (RR 0.75; 95% CI 0.45 - 1.25; p = 0.19). Toxicity or adverse events were described in 18 of the 21 studies. In general, few serious adverse events were reported. Toxicity associated with bisphosphonates is generally mild and infrequent. Renal toxicity is the main issue with intravenous zolendronate and is dose (8 mg) and infusion time related (< 15 minutes). With daily oral calcium (500 mg) and vitamin D (300-400IU) no significant renal impairment or hypocalcamia was observed with a 15 minute infusion of 4 mg IV zolendronate compared with 90 mg pamidronate. Monitoring of renal function with every cycle of zolendronate was undertaken in all studies and is recommended in practice. No significant renal toxicity was observed with intravenous pamidronate or ibandronate. Mild gastrointestinal toxicity is the main toxicity with oral clodronate and oral ibandronate.

Authors' conclusions: In women with advanced breast cancer and clinically evident bone metastases, the use of bisphosphonates (oral or intravenous) in addition to hormone therapy or chemotherapy, when compared with placebo or no bisphosphonates, reduces the risk of developing a skeletal event and the skeletal event rate, as well as increasing the time to skeletal event. Some bisphosphonates may also reduce bone pain in women with advanced breast cancer and clinically evident bone metastases and may improve global quality of life. The optimal timing of initiation of bisphosphonate therapy and duration of treatment is uncertain. In women with early breast cancer the effectiveness of bisphosphonates remains an open question for research.