Detection of recurrent genomic alterations provides new prognostic biomarkers, enables selection of patient groups that may most benefit from specific targeted agents, predicts their response to targeted therapy, and affords the opportunity to elucidate both intrinsic, tissue-specific and acquired resistance mechanisms. With the advent of personalized medicine in cancer, the need for comprehensive genomic profiling of difficult-to-treat tumors is becoming more apparent. While a wealth of information is being generated in the process, characterizing biomarkers for patient classification, prognosis, predicting drug response, and resistance to treatment is crucial.
Defining prognostic and predictive biomarkers in breast cancer is more complicated than in other tumor types. This is primarily because breast cancer represents a heterogeneous set of diseases with distinct molecular features, natural course of disease, and response to treatment. Recognition of this heterogeneity in more recent studies has allowed more precise understanding of molecular characteristics that influence drug response and patient outcomes. Breast cancers are clinically subtyped based on three biomarkers: expression of estrogen receptor (ER) and progesterone receptor (PR) as assayed by immunohistochemistry (IHC) and expression of human epidermal growth factor receptor 2 (HER2) or amplification of erb b2 receptor tyrosine kinase 2 (ERBB2) as assayed by IHC or fluorescence in situ hybridization (FISH), respectively.14,15 In addition to patient stratification, these biomarkers are useful for both prognostication and for predicting response to targeted therapy, eg, ER expression for endocrine therapy;16 ERBB2amplification for trastuzumab.17 The breast cancer subtype that lacks the expression of ER, PR, and HER2 or ERBB2 amplification, commonly referred to as triple-negative breast cancer (TNBC), is associated with poor prognosis, and standard cytotoxic chemotherapy remains the mainstay of treatment. Gene expression profiling assays (such as OncotypeDx®, Mammaprint®) can also further stratify ER+ cancers to identify those cancers that may benefit from addition of chemotherapy to standard hormonal therapy. ERBB2amplification is currently the only genomic alteration that is routinely assayed as part of clinical care of breast cancer.
Genomic profiling using high-throughput, next-generation sequencing technologies (ie, whole-exome, whole-genome, and RNA-seq) has identified recurrent point mutations in breast cancer subtypes: PIK3CAmutations in ER+ breast cancers as opposed to TP53, PTEN, and BRCA1 mutations in TNBCs.18–20Tumors associated with germline BRCA1 mutations can be specifically targeted, either with PARP inhibitors21 or with platinum agents.22 Multiple PIK3CA inhibitors are also in development, currently in both early and late phase clinical trials. Unlike other genomic alterations, the relevance of fusions and rearrangements in breast cancer and its treatment has been less well described.
In this review, the current knowledge of chromosomal instability (CIN) in breast cancer and implications of prognosis for the different molecular subtypes is explored. The patterns and frequency of genomic rearrangements in breast cancer are also discussed. Since more recent knowledge on genomic rearrangements relies heavily on the technique used to study them, the most relevant roles of different technologies and the information acquired are described. Given the therapeutic potential of fusions in cancer, the benefits of identifying and characterizing such rearrangements in breast cancer is outlined. Finally, the current knowledge of breast cancer-related fusions as predictive biomarkers, the future of this evolving field, and the clinical potential for improving therapeutic options for patients with breast cancer is discussed.
Chromosomal Instability and Breast Cancer Prognosis
Genomic rearrangements are closely associated with CIN, which is defined as a dynamic state in which gains or losses of whole or parts of chromosomes occur. Such instability can alter the number of chromosomes, a phenomenon known as aneuploidy. Some cancers such as breast and colorectal cancers harbor more CIN as compared to others.23–25 Though the mechanism of CIN is poorly understood, the implications of its extent have been investigated in relation to clinical outcomes in breast cancer subtypes.
Several groups have analyzed the overall patterns of CIN of breast cancer subtypes and the relevance to clinical outcomes (Table 1). In these studies, CIN was measured based on DNA copy number changes and losses or gains of chromosomal regions or chromosomal number changes in tumor nuclei.26–28 To evaluate the prognosis in some cases, data were retrospectively compared to the clinicopathological details of treatment-naïve patients. CIN is generally considered to be associated with poor prognosis in solid tumors.29,30 While analysis between ER+ and ER− subtypes confirmed that this is true for ER+ cancers, extreme CIN did not show a clear association with survival outcome or prognosis in ER− cancers.26,31Another study evaluating CIN on the basis of copy number changes found a correlation between increased CIN and poor survival outcomes in ER+ and HER2+ subtypes.27 Similar to the previous reports, CIN score was higher in ER− and TNBC samples. However, there was no correlation with survival outcome in these tumors. Major differences between the studies, including small sample size within CIN study cohorts, different methods for evaluating CIN, presence of confounders, absence of detailed patient and treatment profiles, and other parameters, make interpretation and comparison of such data difficult. These studies highlight the complexity of breast cancer genomes and point to the fact that while instability might be used to assess risk in ER+ cases, additional distinct biological markers for predicting clinical outcome in ER− cases are needed.
(To view a larger version of Table 1, click here.)