Abstract IA13: Extrinsic control of the breast cancer spectrum

Abstract

Abstract Breast cancer consists of multiple subtypes characterized by genetic, transcriptional, histologic, and clinical differences that confer distinct prognoses. While epidemiology associates distinct classes with certain groups of women, it is unclear what aspects of biology determine the spectrum of breast cancer observed in different populations. Ionizing radiation is a well-established human carcinogen, particularly of breast. Of particular concern is that girls who have been successfully treated with radiation for childhood cancers have a relative risk of subsequently developing breast cancer similar to that conferred by BRCA1 germline mutation. Moreover, recent studies specifically found that the invasive breast cancer arising in previously irradiated breast tissue is more likely to be triple-negative (TNBC, i.e., negative for hormone receptors and HER2 amplification) compared with age-matched sporadic invasive breast cancer (1). Cancer risk following radiation exposure is often considered exclusively through the prism of the cell-intrinsic effects of DNA damage and mutations that might initiate transformation in long-lived cells. For example, a model proposed by Vogelstein and Tomasetti attributes cancer risk to the frequency of tissue stem cells and the 3 factors that increase their mutation load, hereditary defects, environmental mutagens, and replication; the latter that occur stochastically have been dubbed the bad luck mutations (2). Although radiation is a mutagen, the cell-intrinsic paradigm cannot explain the high rate of TNBC in radiation-preceded breast cancer. Over the last decade, research has “zoomed out” to not only include intrinsic genetic and epigenetic alterations of cancer cells, but to also incorporate tumor microenvironment constituents as essential partners in cancer development (3). We postulated that the effect of ionizing radiation on tumor type and age dependence could act through local and systemic effects. To test how host biology shapes the course of carcinogenesis, we used a genetic-mammary chimera in which Trp53 null epithelium is transplanted to wild-type syngeneic hosts. The Trp53 null epithelium gives rise to an apparently normal ductal outgrowth with a high rate of malignant transformation that generates carcinomas over the course of 12-18 months. These murine mammary cancers are biologically and genomically diverse in a manner very like that of human breast cancer (4, 5). To assess the contribution of radiation effects on stromal-epithelial-immune interactions during carcinogenesis, we irradiated mice prior to Trp53 fragment transplantation (i.e., a radiation-genetic chimera). These experiments show that radiation exposure acts via the microenvironment to promote fast-growing, estrogen receptor (ER)-negative cancers (6, 7). Notably, the signature of Trp53 null tumors arising in irradiated mice clusters cancers obtained from irradiated people and sporadic TNBC (4). Our recent focus is the basis for the increased growth rate of tumors arising in irradiated mice. New studies show that aggressive, rapidly growing ER-negative tumors arising in irradiated mice are enriched in immunosuppressive myeloid cells and depleted of tumor-infiltrating cytotoxic T cells compared to matched cancers arising in nonirradiated mice. Treating irradiated mice with a nontoxic compound that reduces inflammation eliminates these effects of host irradiation but has no evident effect on nonirradiated mice, suggesting that low-grade chronic inflammation may be a target for preventing or delaying breast cancer in women post-irradiation. Our extensive transcriptomic and histologic characterization of Trp53 null mammary cancers show that the distribution of subtypes is modulated by host genetic background, prior radiation exposure, type of radiation, and age. These studies lead us to propose that the breast cancer spectrum results from dominant extrinsic processes that likely influence selection of sporadic genomic mutations.