To identify large numbers of antigens that can be used to recognize the presence of cancer by detecting antibodies to tumor proteins in the serum of the test subjects. Our technology will provide an early detection test for breast cancer in asymptomatic women. We will use bioinformatics techniques to analyze these protein microarray-immunoassays to discriminate between cancer patients and healthy subjects so as to detect disease prior to standard diagnoses as well as discriminate patients with benign conditions or other cancers that might be a false positive in less specific assays.
The aims are organized in 4 subgroups: Aims 1-4 will identify the more informative antigen biomarkers for use on our microarrays. Aim 5 will identify the protein sequences in these informative cDNA clones. Aims 6-8 will provide validation on larger serum sample sets. Aims 9-10 will establish a format of the diagnostic suitable for collaboration with other EDRN components.
Aim 1. Employ high throughput antigen cloning and identification of a suitable number of biomarkers to accurately identify invasive breast cancer.
Aim 2. Use machine-learning techniques such as classification trees and neural networks to develop classifiers able to distinguish women with breast cancer from healthy subjects.
Aim 3. Estimate the test characteristics (sensitivity, specificity, accuracy) of the antigen markers for distinguishing sera of breast cancer patients from control sera.
Aim 4. Eliminate those antigen markers that cross-react with patients with nonmalignant breast diseases and autoimmune diseases.
Aim 5. Determine the DNA sequence of the cDNA clones of validated markers.
Aim 6. Validate the markers on larger sample sets. We will use antigen clones to detect serum antibodies in the BCPS serum set. We will perform a secondary analysis of the reaction of antibodies in the patients’ sera to identify whether there are antigen reactions indicative of a high likelihood of recurrence.
Aim 7. Determine the accuracy of the test in sera prospectively derived from women who later developed breast cancer.
Aim 8. Eliminate antigen clones that react with sera from people with other cancers.
Aim 9. Prepare of purified antigen-containing proteins for the penultimate antigen diagnostic array.
Aim 10. Perform internal and external validation tests with EDRN partners.
We are using IgG molecules specific to cancer patients as a bait to clone antigens derived from phage T7 display cDNA libraries of mRNA from tumor tissue that we use as biomarkers for the early detection of cancer. In a sense we are using the immune system as a biosensor. The immune system elaborates antibodies which we detect as they react with the antigens we clone. The antigens are expressed and then robotically spotted on microarrays. The microarrays are treated with sera from other patients and the binding of IgGs in those sera is used to find the most commonly reactive antigens. We print replicates of thousands of antigens on the microarrays and analyze them with a two-color detection system. The patients' IgGs bound to the spotted antigens are detected using a secondary antibody against human IgG labeled with Alexa-647, a red dye. The second color provides a control for each spot. We use a monoclonal antibody to the N-terminal 11 amino acids of phage backbone protein onto which each antigen is cloned as a fusion protein. That monoclonal antibody is detected with an Alexa-532 (green dye) labeled antibody against murine IgG. The dye ratios provide a control for variations in spotting so as to quantitate the IgG binding to each clone using sera of each patient. We also have a negative control clone that contains no additional amino acids in the cloning vector. The dye ratios on the chips are normalized and those data used in a t-test using sera from patients and healthy controls. Those clones significant in a t-test are used on a validation set of patients and controls not including the previous training samples. The antibody binding data of those subjects in the validation set are tested using machine learning techniques such as neural networks and n-fold cross validation to determine the accuracy of the classifiers.
There are currently no biomarkers annotated for this protocol.
No datasets are currently associated with this protocol.
The National Cancer Institute's Division of Cancer Prevention has released a new funding opportunity to solicit organ-specific applications for Biomarker Developmental Laboratories (BDLs), one of the four scientific units of the recently funded Early Detection Research Network (EDRN). The EDRN is a national infrastructure funded to discover, develop, and validate biomarkers for risk assessment, detection, and molecular diagnosis and prognosis of early cancer. BDLs are responsible for the discovery, development, characterization, and testing of new, or the refinement of existing, biomarkers and biomarker assays for risk assessment, detection, and molecular diagnosis and prognosis of cancers.
The existing BDLs are primarily focused on ovary and gastrointestinal cancers. The proposed BDLs (to be supported under this funding opportunity) should be focused on one or more of the following cancers: breast, prostate and other genitourinary organs, or lung. In addition, cancers with rapidly rising incidence rates, e.g., endometrial, hepatocellular, kidney, thyroid, oropharyngeal cancers, and/or cancers with unique etiology, e.g., mesothelioma, will be considered.
The newly funded units of the Early Detection Research Network will be announced later in April. Successful applicants have already been notified. Those researchers who were not successful during the last round of applications are encouraged to apply to this opportunity.