Breast cancer is the most common cancer in Austrian women. Estimation of prognosis and treatment strategies is increasingly being dependent on stratification of tumors into different entities or classes. Currently, clinical routine stratification of tumors is mostly based on hormone receptor, HER2 status, and estimation of proliferation. However, a more robust and objective classification of tumors can be achieved by elucidation of further biological properties, which is also of increasing significance, as novel anticancer therapies are based on biological mechanisms. Consequently, available information from molecular analyses is increasingly being implemented in routine diagnostic assays with the aim to improve stratification for optimal treatment selection. To date the most extensive molecular-based taxonomy of breast cancer has been achieved by a classification based on combining gene expression and somatic copy number alterations (SCNAs), referred to as integrative clusters. Tissue biopsies are the current gold standard to attain such a classification. However, they can often be difficult to obtain in the metastatic setting and are subject to sampling bias due to intratumor heterogeneity. "Liquid biopsies" are, among other analytes, based on the analysis of cell-free DNA (cfDNA) which contains circulating tumor DNA (ctDNA), i.e. DNA fragments shed from normal and tumor cells into the blood, in patients with cancer. cfDNA can be obtained minimally invasive with a blood draw, allows for the "real time" analysis of tumor DNA from the circulation, and blood samples can be repeated at any time point, which is especially important for monitoring response to therapy. The investigator's group has extensive expertise in the analysis of cfDNA and has developed a plethora of approaches for ctDNA analysis. Recently, the investigators have developed a new approach, which relates to nucleosome positions and gene expression. cfDNA fragments have been associated with the release of DNA from apoptotic cells after enzymatic processing and hence consist mainly of mono-nucleosomal DNA. By performing whole-genome sequencing of cfDNA the investigators could demonstrate that at transcriptional start sites, the nucleosome occupancy results in different read-depth coverage patterns in expressed and silent genes. By employing machine learning for gene classification, the investigators were able to classify genes in cells releasing their DNA into the circulation as expressed. The main hypothesis of the project is that integrative breast cancer clusters can be established from directly blood without the need for an invasive tissue biopsy. Hence, the study aims include refining stratification of patients for an improved selection of treatment strategies. Furthermore, the investigators will obtain novel insights into the biology of metastatic breast cancer, so that this project will have important implications for patients, clinical oncologists, pathologists, pharmacologists, and all basic researchers interested in cancer.

Plasma DNA and DNA from primary tumors