Mammography reduces breast cancer mortality, but it does not always detect cancer early enough to avoid invasive procedures such as nodal dissection, radiation, and chemotherapy. Our vision is that a simple blood test could be used in conjunction with mammography to detect all breast cancer early in the disease process. For example, one clinical scenario might work as follows: at the time of her annual physician visit and mammogram, a woman could have her blood drawn and tested for a panel of breast cancer markers. If the mammogram were clearly positive, she would be referred for biopsy. If the mammogram were equivocal, she would be referred to biopsy only if the marker panel suggested a malignancy. If the mammogram were negative, but the marker panel suggested malignancy, she would be referred for additional imaging such as magnetic resonance imaging (MRI). Other scenarios may be preferable, and indeed, one of the purposes of our study is to work with members of the advocacy community to explore other possibilities.

Our goal is to evaluate a variety of breast cancer biomarkers for their contribution to early detection of breast cancer. We are particularly interested in markers that identify aggressive forms of breast cancer that are missed by mammography. We will work with laboratory scientists to improve the markers so that they predict breast cancer accurately. Then, we will select the best markers for inclusion in a marker panel and evaluate its performance when used with mammography to detect breast cancer early. Finally, we will analyze matched blood samples and tissue from women with breast cancer to confirm that the markers in the panel identify the most aggressive forms of breast cancer.

We will assemble the best team of scientists to accelerate progress in early detection of breast cancer using markers that can be measured in blood. We will develop a unique resource for inter-institutional collaborative breast biomarker research and use it to evaluate the performance of candidate breast cancer biomarkers in women participating in mammography. We will define how best to use this panel of biomarkers together with mammography to detect all breast cancers before they spread locally or to distant sites.

We will develop a specimen resource from a well-characterized population, with associated risk factor information, mammography findings, and follow-up data on cancer outcomes, including blood samples from selected women who have mammograms, biopsy, or surgery. We will evaluate candidate biomarkers for their ability to distinguish among women with healthy breasts and women with various breast conditions, including invasive ductal and lobular carcinoma, LCIS, comedo- and noncomedo-type DCIS, hyperplasia and other potentially premalignant conditions, and benign conditions. We will evaluate the role of biomarkers in improving our current breast cancer early detection strategies.

We will develop state-of-the-art information systems to support patient and specimen tracking, specimen inventory control, and analysis of data, including a web-based interface and a user-friendly query capability. Communication and collaboration will be facilitated via a web-based knowledge management system. Rather than being message-based like e-mail, it is idea/entity-based. Each topic or idea is given its own web page within the system. Collaborators can add their comments and discussion to any of these topics. In addition to this, certain research ¿entities¿ like manuscripts, summary statistics, journal references, timelines, protocols, common data elements, meeting minutes, etc. can be posted and referred to from any of the discussion topics. Finally, all of these entities can be interlinked via hyper-text links allowing the reader to jump quickly between relevant discussion, data, and references.

A highly qualified team of advocates will work with us to ensure the clinical relevance of the marker panel. Markers could be ordered before, at the same time as, or after a mammogram. Results could be used by the primary care physician, radiologist, or specialist who is asked to evaluate suspicious findings. Each strategy will affect the performance of the combination of tests, including mammography. If markers identify cancers that cannot be seen on a mammogram, clinical work-up to identify the location of the tumor will be required. Cost-effective strategies will be needed. Because we will obtain blood samples for women in several stages of evaluation¿prior to screening, just prior to biopsy, and just prior to surgery¿we will be in a unique position to evaluate the clinical utility of markers at all these potential stages of screening. We can then clearly outline the tradeoffs in terms of accuracy, false positives, and costs of detection for alternative strategies.