Institutt for biomedisin

BBB seminar: Bodour Salhia

Development of better predictive, preventative and curative treatments for metastatic breast cancer


Bodour Salhia
Integrated Cancer Genomics Division, Translational Genomics Research Institute, Phoenix, AZ, USA

Cancer metastases arise from disseminated cells of the primary tumor mass before treatment and/or from minimal residual disease (MRD) persisting after therapy (collectively known as micrometastatic residual disease). The ability to predict which patients will develop distant disease recurrence is still based on relatively crude factors. A number of clinico-pathological criteria and molecular tests like MammaPrint and OncotypeDX have been established as breast cancer prognostic markers and used to determine risk of recurrence and stratify patients into high and low risk groups. Even with these clinico-pathologic criteria, we are still unable to concretely define which groups of patients will be cured or will develop metastatic breast cancer (MBC) regardless of whether they are stratified as having high-risk or low-risk disease. It would be advantageous to predict the subset of patients where the disease may recur and develop trials, which would evaluate the use of additional therapy for eventual prevention of metastasis. Human blood is easily accessible for sampling and contains informational cues from tumors, which “leak” protein and DNA into circulation. In the last few years, circulating cell-free (cf)DNA has attracted attention for clinical use in the context of risk prediction, prognostication and prediction of response to chemotherapy in human cancer. To this end our lab has recently performed an unbiased whole-genome approach for describing the circulating methylome of MBC. Unique to our design was the comparison of MBC to both healthy (H) individuals, and disease free survivors (DFS). Through these comparisons, we have identified a novel 21-gene signature of hypermethylated CpG island (CPGI) hotspots unique to MBC, which we believe can be translated into a future blood test to predict breast cancer in women with early stage breast cancer, bypassing the need for tissue.

Treatment of MBC depends on location of the metastatic tumors and includes surgery, radiation, chemotherapy, biological, and hormonal therapy. However, prognosis remains poor and most patients are refractory to therapy. In particular, recurrences in the brain have become particularly more common as greater control over systemic breast cancer is achieved. Despite the increasing number of women dying from brain metastasis (BM), most clinical trials exclude these patients due to poor outcomes and blood-brain barrier (BBB) penetration difficulties. Progress in treating MBC in general has been hampered by a lack of clinically reproducible model systems, a lack of access/availability to human tissue samples, and a lack of comprehensive molecular portraits reflecting the heterogeneous, multi-site nature of the disease. Current models of metastasis are principally derived from cell lines and do not represent the biological underpinnings of the human disease. In order to shed light on the molecular biology of these highly heterogeneous tumors and establish resources for preclinical testing we have created both in vivo and in vitro models of CNS metastases originating from lung and breast cancer. We have performed comprehensive molecular analyses of the patient tissues and patient-derived xenograft (PDX) tissues and demonstrated a close resemblance of the models to the human disease. Spurred by results from a series of ongoing studies, we have now generated (in vivo, ex vivo, in silico) data pointing towards the potential efficacy of targeted therapies in brain metastasis which we wish to deploy as future clinical trials.

Chairperson: Frits Alan Thorsen <frits.thorsen@uib.no>, Department of Biomedicine