The impact of tumor stroma on drug response in breast cancer

Abstract

In the last two decades the breast cancer mortality rate has steadily declined, in part, due to the availability of better treatment options. However, drug resistance still remains a major challenge. Resistance can be an inherent feature of breast cancer cells, but can also arise from the tumor microenvironment. This review aims to focus on the modulatory effect of the tumor microenvironment on the differing response of breast cancer subtypes to targeted drugs and chemotherapy.

Introduction

Breast cancer is the most frequent cancer in women and accounts for 14% of cancer deaths in the United States annually [1]. Due to early detection by mammographic screening and also in part due to improved treatments we are beginning to see an impact on breast cancer outcome. Since 1990, mortality rates have decreased 2.2% each year, while incidence rates fluctuated up until 2003 and remained relatively stable since. Additionally, reductions in breast cancer mortality have also been observed in European studies [2].

Breast cancer is a heterogenous disease. The classical, and still routine way, to subtype breast cancer is by immunohistochemical analysis. Based on this analysis, breast cancer can be divided into three main groups: (1) estrogen receptor (ER)α-positive, (2) human epidermal growth factor receptor (Her)2 positive and (3) triple-negative (ERα-negative, progesterone receptor-negative, Her2-negative) subtypes. These subtypes are of clinical relevance in terms of prognosis and targeted treatments against ERα and Her2 are routinely used in two of the three subtypes [3], [4]. With the availability of molecular tools, breast cancer has also been subtyped with greater detail by gene expression profiling. In a seminal paper, Perou and co-workers showed that breast cancer subtypes can be distinguished by clusters of genes that are predominantly expressed in luminal breast epithelial cells, myoepithelial (basal) cells and cells whose expression is associated with elevated Her2 levels [5]. These subtypes are called luminal-like, basal-like and Her2-enriched, respectively. The molecular subtypes overlap with the immunohistochemical subtypes, such that most luminal-like tumors are ERα-positive, basal-like tumors triple-negative and Her2-enriched Her2-positive and vice versa [6]. However, the gene expression portraits allowed greater distinction in tumor subtypes within the three molecular subtypes leading to the identification of a luminal B as well as a luminal A subtype within the luminal-like tumors [7]. The gene expression subtypes are also of clinical importance, as patients with luminal B tumors have a worse prognosis than those with luminal A disease [8]. Later, an additional molecular subtype, the claudin^low subtype, was characterized by a mesenchymal/cancer stem cell-like gene expression profile [9]. Hence, the use of molecular tools allows an even more precise breast cancer subtyping [10], which puts us well on the path to personalized therapy by targeting the key factors to which individual cancer (cells) are addicted [11].

Chemotherapeutic drugs are commonly used to treat breast cancer. These drugs interfere with one of the key hallmarks of cancer, proliferation [12]. Chemotherapeutic drugs either damage DNA and thus prevent DNA synthesis or inhibit genes involved in proliferation [13]. Anthracyclines, such as epirubicin and doxorubicin, combined with fluorouracil and cyclophosphamide are effective treatment options for patients with early breast cancer [14]. Taxanes (paclitaxel or docetaxel) are also often used in Her2-positive breast cancers [15]. In triple-negative disease, treatments with taxane or anthracycline alone or in combination are common [16].

Targeted drugs have also been available for many years. For example, ERα-dependent breast cancers can be successfully treated by anti-estrogens, such as tamoxifen or fulvestrant, or aromatase inhibitors, such as letrozole, in a setting called endocrine therapy [3]. Interestingly, between 1992 and 2008, a shift in the incidence toward ERα-positive tumors has been observed in the United States [1]. This may additionally account for the decline in mortality rate since ERα-positive tumors are associated with better prognosis and targeted endocrine treatments are available. To specifically target Her2-positive breast cancer, very effective drugs are available, such as the anti-Her2 antibodies trastuzumab and pertuzumab, or Her2 kinase inhibitors, such as lapatinib [4]. These Her2 targeted drugs are often combined with taxanes for optimal treatment [15]. Triple-negative breast cancers are still only treated with chemotherapy, as no specific targeted drugs can currently be recommended. Drugs against potential targets, such as epidermal growth factor receptor (EGFR) or poly(adenosine diphosphate ribose) polymerase (PARP), have shown limited success in this subtype [17], [18]. Recently, further analysis of the triple negative subtype revealed additional heterogeneity indicating subsets of triple negative disease that may be more sensitive to specific drugs, such as the luminal androgen receptor (LAR) triple negative subtype which has been shown to respond to androgen receptor antagonists [19].

Despite successes in developing powerful anti-cancer drugs, drug resistance remains a major obstacle in cancer treatment. Drug resistance may be a pre-existing, inherent feature of cancer cells (intrinsic drug resistance) or may be acquired during the course of treatment leading to a relapse after months or years (acquired resistance). Intrinsic or acquired drug resistance is well established on a genetic and/or epigenetic level resulting in a blockage of pro-apoptotic pathways, in overexpression of anti-apoptotic proteins or in more efficient DNA damage repair [20]. Multidrug resistance can also be found through the promotion of drug efflux.

Even for specific drugs, such as anti-estrogens or anti-Her2 antibodies, many different mechanisms have been described that can lead to resistance [21], [22], [23]. For example, Her2-positive breast cancer cells can acquire resistance by mechanisms ranging as different as downregulation of phosphatase and tensin homologue (PTEN), an inhibitor of the phosphatidylinositol-3-kinase (PI3K) pathway [24] through to activation of integrin β1-dependent signaling [25]. Likewise, in the same breast cancer cell line (MCF-7), resistance against the anti-estrogen tamoxifen can be achieved by overexpression of the androgen receptor [26] or by fostering the crosstalk between the G-protein coupled receptor (GPR)30 with EGFR [27]. The complexity of these signaling pathways and their ability to crosstalk with other pathways may explain the wide range of mechanisms that can lead to resistance. For example, ERα shows ligand-dependent and ligand-independent activities and can act as a transcription factor in a genomic fashion or as component of signaling cascades in a non-genomic fashion. Endocrine resistance may be accompanied by loss of ERα and activation of an alternative survival pathway or, more often, by a switch from estrogen-dependent to estrogen-independent ERα-driven survival [22], [23]. Despite these different mechanisms, endocrine resistance is often accompanied by increased activity of the PI3K pathway [28], which can be targeted in variety of ways [29]. One possibility is to inhibit the mammalian target of rapamycin (mTOR), an important effector of the PI3K pathway, and shown to be involved in critical cellular functions, such as cell growth, proliferation and survival [30]. In a recent study it was shown that breast cancer patients that developed endocrine resistance responded positively to the rapamycin homologue RAD001 [31] suggesting that blocking the PI3K survival pathway is a potential therapy option to overcome endocrine resistance in patients [28].

Most cancers are heterogeneous and within a single tumor different subpopulations of cancer cells can be found. In theory, the higher the number of subpopulations present in a tumor, the higher the chance that one will be drug resistant (Fig. 1) leading to a pool of surviving cells called minimal residual disease. Tumor cell heterogeneity therefore confers tumor robustness allowing it to withstand drug treatment, even though only a few cells may survive.

There are two alternative hypotheses that could explain tumor heterogeneity. One hypothesis claims that heterogeneity is caused by clonal evolution, where different cancer cell subpopulations emerge from a single clone in the course of tumor progression [32]. Alternatively, it has also been proposed that cancer stem cells (CSCs) are responsible for tumor heterogeneity [33], [34]. CSCs are cells that give rise to heterogenous progenies and are present in small numbers in primary breast cancers and breast cancer cell lines [35]. Both concepts (clonal evolution and CSCs) may not be mutually exclusive as different subpopulations of CSCs may exist that generate different progenies (CSC heterogeneity) [36], [37]. Supporting the concept of CSC heterogeneity and its impact on the tumor cell phenotype a single mutation in the breast cancer-1 (BRCA-1) gene has been shown to alter breast cancer progenitor cell fate [38].

The concept of CSCs is an attractive model to explain intrinsic drug resistance [39]. Through overexpressing ATP binding cassette (ABC) drug transporter proteins, CSCs are able to export many drugs and are therefore considered to be multi-drug resistant cells [35]. In addition, there are many more features, such as higher expression of anti-apoptotic proteins or a more effective DNA damaging repair system, that render CSCs drug-resistant [40]. Also, as CSCs differ from non-CSC cancer cells in many ways, CSCs may be dependent upon other signaling pathways than their non-CSC neighbors. For example, in ERα-positive breast cancer, CSCs show much lower ERα-levels than bulk tumor and, at the same time, display an expression pattern indicative of a hyperactive PI3K pathway [41]. Hence, in theory, endocrine therapy would be unlikely to affect CSCs and could lead to drug resistance. Nevertheless, there is evidence that CSCs in ERα-positive breast cancers may need ERα-positive non-CSC neighbor cells for activity [42]. This is in line with the concept of stem cell niches which provide an environment that supports stem cells to maintain their stemness [43]. Irrespective of the potential dependency of CSCs on the bulk tumor, CSCs are considered to be a major risk for relapse. Hence, CSC-specific drugs in combination with commonly used drugs that target the bulk tumor may help to improve the outcome of cancer patients [44].