Considerable knowledge has been gained in understanding the biology of breast cancer and identifying biomarkers that can be used in detecting breast cancer, but translation of that knowledge to the clinical setting has been challenging [1]. Clinical validation is the main hurdle in the process. In a case-control study of the Prospect-EPIC (European Prospective Investigation into Cancer and Nutrition) study in which more than 300 breast cancer patients and matched controls were tested for breast cancer over a period of 3 years using a panel of eight serum markers (osteopontin, haptoglobin, cancer antigen 15-3, carcinoembryonic antigen, cancer antigen-125, prolactin, cancer antigen 19-9, and alpha-fetoprotein), very low specificity (50 %) and sensitivity (50 %) were observed [112]. This may be due to the presence of different breast cancer subtypes in collected samples. Such epidemiologic studies should select a broader target set of potential biomarkers that could be enabled by antibody array technologies in which profiles of up to 100 antibodies can be followed simultaneously. Making different groupings based on the status of hormone receptors (estrogen and progesterone) also might be helpful.

The need to identify and characterize early cancer diagnostic biomarkers is considerable because cancer is a heterogeneous disease, and a patient’s individual molecular profile, resulting from tumor microenvironment, determines disease development and response to treatment [37, 113]. Tumor microenvironment is affected by several factors, including epigenetic factors of the cell. As noted above, a number of noninvasive biomarkers for early detection of breast cancers have been identified.

In the field of cancer prevention, the National Cancer Institute (NCI) has an ongoing study, “the Study of Tamoxifen and Raloxifene (STAR),” with the objective of examining how the drug raloxifene compares with the drug tamoxifen in reducing the incidence of breast cancer in postmenopausal women who are at increased risk of developing the disease. Early results indicated that both drugs are equally effective and that about one-half of the disease incidence could be reduced. After the trial was continued for a longer time in the absence of these drugs to see whether previous prevention could be sustained, a 50 % reduction in incidence was seen with raloxifene, while a 38 % reduction was seen with tamoxifen. Both noninvasive and invasive cancers were studied in the STAR. Furthermore, participants taking raloxifene showed fewer side effects (e.g., uterine cancer, blood clots, cataracts) than those taking tamoxifen. Researchers with the National Surgical Adjuvant Breast and Bowel Project (NSABP) are conducting this study (financially supported by the NCI, USA), and more than 500 participants aged 35 years or older are involved in the study. The trial began in 1999 and stopped enrolling new patients in 2004.

The main areas in which progress and attention are needed with respect to biomarker identification are cost, high throughput, and the application of breast cancer biomarkers in clinical settings. Proper analytical and clinical validation of early markers has not been achieved. Currently, the key challenge in the field is the clinical validation of identified biomarkers. The NCI has developed guidelines for the analytical and clinical validation of biomarkers, but no biomarkers have been validated to date [37, 113–117]. Integrating genomic and proteomic markers with epigenetic markers may be useful in subtyping different cancers and cancer stages [46, 118]. Methylation profiling results from blood and tissues often differ. Koestler and colleagues conducted a systematic epigenome-wide methylation analysis and demonstrated that shifts in leukocyte subpopulations might account for a considerable proportion of variability in these patterns [119]. Multiplexing of biomarkers may reduce false-positive results in screening studies when the intent is to identify populations who are at high risk of developing breast cancer.

Quantitative imaging data storage and maintenance present their own challenges, as discussed above. Whether miR expression is localized in a specific part of the breast tissue must be evaluated carefully. In a tissue biopsy, the local concentration (number of miRs) may be low or high. Determining the accurate level of miR concentration is critical.

Association studies are extremely powerful in identifying new low-penetrance SNPs (biomarkers) that may have therapeutic implications. Identifying common low-susceptibility alleles is useful because it provides possible insight into the mechanisms of tumor biology and identification of high-risk individuals. Because genotyping is no longer expensive, the information from such studies can be utilized in personalized medicine in the form of targeted primary and secondary prevention.

Prognostication is a promising area in the translation of experimental research into clinical practice. In this approach, patterns of altered gene expression in tumors are used to construct classifiers instead of standard indices such as Nottingham Prognostic Index, Adjuvant! Online, and PREDICT [120–122].

In conclusion, we would like to emphasize that considerable progress has been made in identifying breast cancer biomarkers that can be used in the complete spectrum of carcinogenesis, from risk assessment to survival follow-up. The information discussed in this chapter may be useful in developing new interventions and therapeutic targets.