Evaluation of protein and biological biomarkers in breast cancer

Sammanfattning: Breast cancer is the most common malignancy in women around the world. There have been great improvements in treating the disease and today between 80-90% of the women survive at least 5-years after their primary diagnosis. Still, due to the high incidence of the disease more than 450.000 women die of breast cancer each year worldwide. Much of the improvements in breast cancer survival can be explained by better knowledge of the development and progression of the disease, hence the treatments have become more effective. Yet, we still cannot explain why one patient relapses and dies whereas another patient, with seemingly similar tumor, survives. Thus, the identification of novel and evaluation of present breast cancer biomarkers are vital steps in the progression of improving the survival rate of breast cancer patients. Estrogen receptor alpha (ERα) is expressed in around 70% of the tumors. ERα is a good prognostic and predictive biomarker since it can effectively be targeted by endocrine treatment. The gene Dyslexia 1 candidate 1 (DYX1C1) has been shown to be overexpressed in cancer and also to regulate and be regulated by estrogen and its receptors. We evaluated the expression of DYX1C1 mRNA and protein in three independent breast cancer cohorts and its association to known breast cancer biomarkers and survival (study 1). We observed that DYX1C1 expression was positively associated with ERα expression and also functioned as a prognostic marker for improved survival. Biomarker expression in breast cancer is usually examined using immunohistochemistry (IHC) on whole tumor sections. During the diagnostic process, cells from fine-needle aspirations are usually examined, sometimes also by immunocytochemistry (ICC). ICC is also useful in a metastatic setting, and it is therefore important that the results are concurrent with IHC evaluation. We retrospectively examined paired IHC and ICC evaluation of ERα, progesterone receptor (PR) and Ki67 (study 2). We found that there were significant differences in the grading using IHC compared to ICC. Thus, to ensure the proper pathological diagnosis of metastatic lesions, comparisons and validation of these methods to detect biomarker should be performed. Estrogen receptor beta 1 (ERβ1) is expressed in both normal mammary tissue and malignant breast tumors. In vitro data point towards a protective role of ERβ1 with lower proliferation and less invasiveness when overexpressed. However, in vivo data are so far inconclusive, where some previous studies have reported better a prognosis, while others have reported no association or even a worse prognosis. We examined ERα, ERβ1 and splice variant ERβcx in breast cancer patients with clinically negative lymph node status (study 3). We found that ERβ1 was an independent marker of good prognosis. Interestingly it was expressed in patients of all grades and age groups, whereas ERα was commonly expressed in low-grade tumors of older patients. ERβcx did not show any prognostic association, nonetheless, high expression was coupled to risk of synchronous lymph node metastasis. Breast cancer stem cells have been shown to be highly tumorigenic, thus functioning as a biomarker of poor prognosis. The origin of these cells is not completely understood. Either they derive from normal stem cells or they are transformed from differentiated cells of the bulk tumor. Through exome sequencing we found that isolated breast cancer stem cells and mixed cells from the bulk of the tumor did not differ genetically (study 4). We showed that mutations were present in both stem cells and non-stem cells of the bulk tumor and at the same allele frequency. Our data supports a transformation of stem cells from differentiated cells. In conclusion, the evaluation of new and existing cancer biomarker has the potential to generate novel hypothesis on tumor biology and reveal new targets for treatment.