Optimizing breast cancer survival models based on conventional biomarkers and stromal parameters

This thesis deals with multiple aspects of breast cancer risk stratification after locoregional treatment. The first part of the thesis deals with the reproducibility of established pathological parameters that currently stratify breast cancer patients to low- or high risk, on the basis of which systemic therapies are considered. The reproducibility of estrogen receptor, progesterone receptor and HER2 is investigated. Additionally, prognostic implication of lymph vascular space invasion is assessed as well as its interobserver reproduciblity. Lastly, the reproducibility of Ki67 assays are determined. The second part of the thesis concerns investigations into the prognostic aspects of the tumor-associated stromal tissues. These tissues might be used to further improve breast cancer risk stratifications as well as help determine tumor susceptibility to systemic treatments. The prognostic implications of the tumor-stroma ratio is investigated. Proteomic studies into the tumor-associated stroma is described as well as a work-flow for investigating metabolic interactions between the tumor epithelium and tumor stroma. The prognostic significance of TGF-beta signaling is also investigated.