Many of the investigations in this genetic epidemiology project arise from observations in families at high risk of cancer or in other etiologic studies. Case-control study data from 718 non-Hispanic white patients with invasive cutaneous melanoma from melanoma clinics in Philadelphia and San Francisco and 945 matched controls from outpatient clinics with similar catchment areas to the cases were used to develop a risk assessment model for estimating 5 year absolute risk of melanoma. The risk assessment tool is available on the web at www.cancer.gov/melanomarisktool/ . The attributable risks of melanoma using the gender specific models were 86% for men and 89% for women using simple variables that are easily obtainable by health care professionals. Questionnaire data, tumor blocks, and DNA from a case-control study of 183 incident melanoma cases and 179 controls conducted in North-Eastern Italy were used to further evaluate MC1R and BRAF. In subjects with melanoma arising on sun exposed areas of the body and with limited signs of chronic solar damage, we found a strong association between MC1R germline variants and melanoma with somatic mutations in the BRAF oncogene. We are now analyzing other genes involved in melanogenesis and signal transduction pathways to further explore the etiology of melanoma.As a follow-up to a DCEG comprehensive case-control study of adults with brain tumors, a family-based study of the parents, siblings and adult children of the 480 eligible glioma cases has been conducted. Relatives of 365 of the glioma cases were interviewed about personal and family medical history and other risk factors and were asked to provide buccal cells as a source of DNA. Analyses to examine the risk of cancer among the first degree relatives compared to population controls are in process. Examination of questionnaire data from a retrospective single institution study of childhood medulloblastoma showed little evidence for increased cancer risks among relatives of medulloblastoma patients.In collaboration with Yale University Tissue Microarray (TMA) core facility, we have successfully built TMAs of 842 invasive tumors, with two-fold redundancy, collected from a population-based case-control study of breast cancer conducted in Poland. We have immunohistochemically stained these TMAs for 18 molecular markers involved in hormone biosynthesis, metabolism, and receptor mediated pathways. Analyses of molecular signature markers (ER-alpha, PR, HER2, EGFR, and cytokeratin 5) suggest that risk factors for breast cancer may vary by molecular subtypes. We have also stained these arrays for four markers (ER-alpha, ER-beta, PR and HER2) using a newly developed Automated Quantitative Analysis (AQUA). Comparison of the two approaches (AQUA and IHC) for the same four markers demonstrated AQUA analyses of tumors represented in TMAs provide reliable, quantitative measures of marker expression. Currently, a new set of invasive TMAs including 919 Polish breast cancer cases with three-fold redundancy is being constructed at Yale. Once completed, we will have 1,761 invasive tumors with at least two-fold redundancy built on TMAs. We have published results detailing a technically challenging method to construct TMAs from non-invasive epithelial tissues. Using this technique, we have successfully built 32 TMA blocks of 1,547 tissue cores including both non-invasive tissues and their associated invasive tumors (N=357) collected from 560 Polish breast cancer cases. We have stained these non-invasive arrays with ER-alpha, PR, and HER2. We will use these TMAs as platforms to study etiologic heterogeneity of breast cancer characterized by expression patterns of molecular markers in both invasive tumors and their adjacent non-invasive epithelial lesions and stromal tissues.
Showing the most recent 10 out of 29 publications
