Mammostrat was derived by taking advantage of insights from gene expression studies that established that breast cancer patients have many different subtypes of tumors.   Hundreds of antibodies were created and screened for their ability to distinguish different subtypes of breast cancer. For more information on the technology behind the Mammostrat test including AGI's pipeline of testing, go to the Technology Section of the AGI website.

Mammostrat was initially discovered in a study in which AGI applied its breast 'Panel of Diversity' to a 550 patient cohort of breast cancer specimens from the Clearview Cancer Institute of Huntsville, AL. AGI first assessed the association of each antibody individually with clinical outcome in the ER+ and ER+, node negative patient subsets. Statistical analysis was used to nominate prognostic models that minimized the number of biomarkers needed to predict outcome. A set of five antibodies combined using a Cox proportional hazard model was nominated as the favored model. This early version of the test consisted of mainly polyclonal antibodies and was used for the initial clinical validation studies.

Independent testing of this first generation Mammostrat test used two breast cancer cohorts assembled at the University of British Columbia and The Cleveland Clinic Foundation. These studies demonstrated that the test was strongly associated with outcome (JCO 2006 , SABC 2005). A second generation Mammostrat test was then developed which replaced the polyclonal antibodies with monoclonal antibodies. An additional validation study was performed using the monoclonal antibody test in collaboration with the University of Alabama at Birmingham. This study also showed a strong association with outcome in both all estrogen receptor expressing patients as well as the subset of patients who did not have nodal metastases at diagnosis (ASCO 2006).

In addition, a definitive clinical trial validation was performed in collaboration with the National Surgical Breast and Bowel Project (NSABP) using archived samples from the B14 and B20 clinical trials of adjuvant therapy for node-negative breast cancer. These clinical trial specimens were from the seminal trials that established tamoxifen and adjuvant chemotherapy as effective therapies and therefore provided an unbiased, independent validation of Mammostrat testing. The Mammostrat NSABP validation results were presented at the San Antonio Breast Cancer Symposium in December of 2006 and American Society of Clinical Oncology annual meeting in June of 2007.

Go to the Mammostrat section of Publications to read the publication(s) related to the following independent validation studies:

• The University of British Columbia
• The Cleveland Clinic Foundation
• The University of Alabama at Birmingham
• The National Surgical Adjuvant Breast and Bowel Project