Titles and Affiliations
Professor of Medical Statistics
Nuffield Department of Population Health
University of Oxford
Head of the Secretariat for Early Breast Cancer Trialists’ Collaborative Group (EBCTCG)
Research area
Conducting meta-analyses of treatments for early breast cancer to provide the best possible estimates of the balance of benefits and harms of a given therapy.
Impact
The Early Breast Cancer Trialists’ Collaborative Group is a collaboration of over 900 trialists from all groups who have performed randomized trials in early breast cancer. The collaboration brings together all randomized evidence worldwide on the treatment of early breast cancer in meta-analyses –a tool for combining the results of multiple different studies addressing the same question to generate a more precise estimate of an effect—that provide the best possible estimates of the balance of benefits and harms of a given therapy, improving outcomes, and giving women the chance to make informed decisions about their treatment. The proposed work builds upon the existing collaboration and has been identified by the breast cancer community, notably the independent international steering committee, as being the most relevant questions which can be answered at this time.
Progress Thus Far
Over the past year, the Early Breast Cancer Trialists’ Collaborative Group (EBCTCG) prioritized six meta-analyses of treatment, identified as being of importance to clinicians and patients with breast cancer by an independent steering committee formed from leading international experts and patient advocates, and finalized two analyses looking at whether specific tumor characteristics affect response to therapy. This continues the EBCTCG’s aims of identifying successive improvements in early breast cancer treatment, to investigate whether benefits vary by type of tumor, to influence international treatment guidelines and thereby reduce breast cancer mortality worldwide.
What’s next
Over the next year the EBCTCG will prioritize meta-analyses on cardiovascular mortality in 160,000 women in about 100 trials of endocrine therapy, tamoxifen duration, trastuzumab duration, axillary treatment and partial breast radiotherapy, bone-directed therapy, and newer therapies such as CDK 4/6 inhibitors. Other meta-analyses are planned once data become sufficiently mature. Additionally, many biomarkers of treatment efficacy are suggested by reports from single trials, but only individual patient data meta-analyses can provide sufficient statistical power to provide validation. The EBCTCG is uniquely poised to foster the collaborations necessary to make the best use of emerging technologies such as digital pathology and AI to accomplish these critical goals.
Biography
Robert Hills, DPhil is Professor of Medical Statistics in the Nuffield Department of Population Health (NDPH) at the University of Oxford and head of the Secretariat for the Early Breast Cancer Trialists’ Collaborative Group (EBCTCG). Originally a mathematician, he has worked on clinical trials and meta-analyses for the past three decades, first joining the CTSU, now part of Oxford Population Health in 1993. For approximately 2 decades he was statistician for the UK MRC/NCRI trials in acute myeloid leukemia and became head of the Haematology Clinical Trials Unit in Cardiff in 2010. He developed a number of novel clinical trial designs, especially in conditions where outcomes are poor but there are promising candidate therapies. These designs allow an early change from an unpromising new therapy to one with greater potential. He also integrated laboratory and clinical data, and worked on individual patient data meta-analyses, where he demonstrated a survival benefit for gemtuzumab ozogamicin in AML. In 2018, he returned to NDPH as part of the EBCTCG. The EBCTCG was conceived in 1983, and for forty years has periodically sought up-to-date data on each individual woman in randomized trials of treatments for early breast cancer for meta-analysis. This worldwide collaboration involves over 1000 trialists from about 650 trials, giving a database of almost three quarters of a million women and 6 million woman-years of follow-up. Through individual patient data meta-analysis, it continues to demonstrate many moderate but real differences in long-term survival that were not reliably identified in the separate trials.
