Original 2002 Literature Search Strategy
The MEDLINE database was searched from 1976 to August 2002 using disease-specific text words and subject headings (breast, mammary, cancer, carcinoma, neoplasm[s]), treatment-specific terms (diphosphonates, bisphosphonates, clodronate, pamidronate, etidronate, alendronate, ibandronate, zoledronate), and design-specific terms (meta-analysis, randomized controlled trial[s], practice guideline). The searches were not restricted by language. Issue 3 (2002) of the Cochrane Library, conference proceedings from the American Society of Clinical Oncology (ASCO) (1997–2002) and the San Antonio Breast Cancer Symposium (SABCS) (2001), and bibliographies were also searched. The Canadian Medical Association (CMA) Infobase (http://mdm.ca/cpgsnew/cpgs/index.asp), the National Guidelines Clearinghouse (www.guideline.gov) and other Web sites were searched for existing evidence-based practice guidelines.
Updated 2004 Literature Search Strategy
The original literature search was updated using MEDLINE (September 2002 to February 2004), the Cochrane Library (Issue 1, 2004), conference proceedings from the American Society of Clinical Oncology (ASCO) (2003) meeting and the San Antonio Breast Cancer Symposium (SABCS) (2002–2003), and bibliographies. Relevant Web sites were searched for new evidence-based practice guidelines.
Inclusion Criteria
Articles were eligible if they met all of the following criteria:
- They were published reports, or abstracts from the American Society of Clinical Oncology or San Antonio Breast Cancer Symposium meetings.
- They presented results of a meta-analysis or randomized controlled trial that compared:
- Treatment with a bisphosphonate to observation or placebo
- Two bisphosphonates
- Two or more doses of the same bisphosphonate; or
- The same bisphosphonate given by two routes of administration
- Trial participants were primarily patients with breast cancer (early-stage or advanced) although trial participants could also include patients with other solid tumours or myeloma.
- Results were reported, by treatment group, for at least one of the following outcomes: survival, quality of life, and adverse effects. Additional outcomes of interest for patients with bone metastases from breast cancer included bone pain (measured using a pain scale or analgesic consumption) and skeletal events, other than hypercalcemia (as bisphosphonates are acknowledged to be an effective intervention for this complication). The development of bone metastases was also an outcome of interest in patients without bone metastases at the time of randomization.
Evidence-based practice guidelines and systematic reviews addressing the guideline questions were also included.