Systematic Reviews of Treatments for Cancer and Methodological Research

Previous and current research:
The differences in outcome between many treatments for cancer are moderate but potentially very important to the public health. However, most trials have been too small to assess such differences reliably on their own. This can be overcome through large prospective trials and systematic reviews of completed trials. The Cancer Overviews group at the Cancer Studies Unit has been attempting the latter for more than 15 years through collaborative overviews in which patient data from all relevant randomised trials are collected and reanalysed. Michael Clarke, who now works part-time for Cancer Research UK, coordinates this work.

The most well established project is the Early Breast Cancer Trialists' Collaborative Group (EBCTCG) overview of operable breast cancer. It involves data from over 200,000 women and has provided definitive evidence of, for example, the effects of tamoxifen and chemotherapy as adjuvant treatments for breast cancer. Such projects represent the 'gold standard' for reviews, but research is needed to assess the relative importance of the different steps in the review process. This will provide reliable evidence on which to judge the quality of systematic reviews. Michael Clarke is at the forefront of this research as Associate Director (Research) at the UK Cochrane Centre. He is deputy chair of the Cochrane Collaborationís international Steering Group.

Future projects:
The fourth cycle of the EBCTCG overview began in 1999. It includes analyses of different durations of tamoxifen, of anthracycline-based chemotherapy and of high-dose chemotherapy. Preliminary findings were presented to the collaborating trialists in September 2000 and heavily infuenced the NIH Consensus Conference on breast cancer in November 2000. The methodological research at the UK Cochrane Centre investigates bias in research synthesis.