Mechanisms behind context-dependent role of glucocorticoids in breast cancer progression

Breast cancer is the most common cancer in women worldwide with an estimated 2.3 million new cases yearly [1]. Histologic classification is evaluated based on the growth pattern (in situ vs. invasive). The most common type (70–80%) is infiltrating duct carcinomas with no special type (IDC), followed by invasive lobular carcinomas (ILC, ~ 10%) [2, 3]. The remaining ones (mucinous, cribriform, micropapillary, papillary, tubular, medullary, metaplastic, and apocrine carcinomas) can be considered rare. Tumor grade (by assessment of histologic differentiation) and stage (TNM, Tumor size, Nodal status, and distant Metastasis) have also important prognostic roles and they are considered in a number of clinical decisions [2]. Early breast cancer accounts for > 90% of all diagnosed breast cancers and despite the availability of different treatment options, ~ 30% of these patients will develop cancer recurrence/progression [4]. Locally advanced or metastatic breast cancers have a median overall survival of approximately ∼3 years and the 5-year survival is only ∼25% [5].

Immunophenotype determined by immunostaining of estrogen receptor (ER), progesterone receptor (PR), and human epidermal growth factor receptor 2 (HER2) provides the basis of targeted therapy selection. Hormone receptors (ER and PR) are expressed in the great majority (∼75%) of all breast cancers and indicates the responsiveness to hormonal therapy [2]. HER2 overexpression is detectable in approximately 15% of breast cancers, due to gene amplification, and it is associated with a more aggressive clinical course and poor prognosis. It is also considered as an important predictive marker indicating the response to anti-HER2 therapy.

Based on global gene expression profile, breast cancer was classified into four subtypes: luminal A, luminal B, HER2-overexpressing, and basal-like breast cancers [2]. Luminal A subgroup is characterized by an immunophenotype of ER + , PR + , and HER2 − , and the proliferation index Ki-67 is low (< 20%). These are typically low-grade tumors with the best prognosis among all subtypes. Luminal B breast cancers are also ER + ones, PR and HER2 status can be both + / − , but Ki-67 index is higher (> 20%). These tend to be higher grade and have a worse prognosis compared to luminal type A tumors [2]. While patients with both luminal type tumors are likely to benefit from hormonal therapy alone, luminal B tumors may be candidates for additional chemotherapy. HER2 overexpressing (HER2 +) subtypes of breast cancer are ER − and PR − , and they likely to be high grade exhibiting an aggressive clinical course. Nevertheless, due to their HER2 positivity, they respond to anti-HER2-targeted therapy which results in an improved outcome. The basal-like tumors (also referred as triple-negative breast cancer showing negative staining of all ER − , PR − , and HER2 −) are characterized by high grade and high proliferation index.

These subgroups usually show good correlation with the immunophenotypic classification, which highlights the significance of the pathological assessment. Indeed, in routine clinical practice, the application of gene expression profile-based classification is limited due to cost and technical challenges [2].

GCs, especially dexamethasone (dex), are routinely administered as adjuvant therapy to manage the side effects of cytotoxic chemotherapy due to their antiemetic effects, energy and orexigenic properties, and to prevent hypersensitivity reactions. However, recently, it has been revealed that dex triggers different effects depending on the breast cancer molecular subtype, which has raised new concerns regarding the generalized use of GC, and suggests that the context-dependent effects of GCs and GC resistance can be taken into potential consideration during treatment design [6-8].

Glucocorticoids (GCs) are steroid hormones synthesized in the adrenal cortex and controlled by the hypothalamic–pituitary–adrenal (HPA) axis. They are classical stress hormones induced by several factors, including psychological stress, infection, inflammation, trauma, toxins, and others. They regulate several cellular processes including metabolism (glucose, protein, lipid, and carbohydrate), immune and inflammatory responses, as well as vascular tone. GCs also widely influence the operation of the central nervous system (arousal, cognition, mood, and sleep), and they have a role in regulation of the circadian rhythm, cell cycle, and programmed cell death [6].

These pleiotropic effects of GCs are involved in the regulation of development of breast tissue, and also in fine tuning of physiological and pathophysiological processes of breast tissue, including breast cancer development (Fig. 1). The dual role of GCs in breast carcinogenesis has been shown in tumor development, cell proliferation, and apoptosis [9]. Chronic stress, disturbed circadian rhythm, inflammation, metabolic syndrome, obesity, depressive behavior, and major depression which have been associated with glucocorticoid effects all have been also linked to increased breast cancer risk and progression [ 7‐15].

All these effects should be placed in certain context because the glucocorticoid effects are highly dynamic both in time and space. Related to tumor and metastasis development, the following points especially emphasize the role of GC action in breast cancer: (1) GCs seem to act as both tumor suppressor and tumor promoter in breast tissue in a cell-type specific manner and in a context-dependent way; (2) in the majority of breast cancer cases, tumor development and growth are initially steroid hormone dependent and involve GR crosstalk with other hormone receptors; (3) GCs are routinely administered as adjuvant treatment for the side effects of conventional chemotherapy.

In this review, following a brief description of the molecular background of the diverse biological effects of the GR, we summarize the already known and potential factors and mechanisms (Fig. 2) that regulate glucocorticoid action involved in breast carcinogenesis, and discuss the potential therapeutic implications.

The GR is expressed almost ubiquitously among human tissues due to the presence of multiple transcription factor-binding sites in its promoter [10]. The detailed structure and action of the glucocorticoid receptor (GR) have been described in excellent reviews [10‐12] and is beyond of the scope of this work; however, a brief summary is essential to the understanding of the diverse functions of GR.

The mechanisms behind different GC responses have been first investigated in conditions where GCs are frequently used, such as chronic inflammatory diseases and hematological malignancies [47, 48]. Different escape mechanisms have been reported and extensively reviewed by, e.g., Scheijen et al. 2019, Ramamoorthy & Cidlowski 2013, and Ciato et al. 2020 [35, 49, 50]. Due to the strong cell type and context dependency of GCs, regarding breast cancer, the potential role of GC usage has been recently challenged by experimental data [7]. Therefore, in the following paragraphs, the mechanisms of altered GC response and context-dependency are summarized (Fig. 2).

The tissue-specific expression of GR itself, its coregulators, and transcription factors result in distinct responses among multiple pathways targeted by a given transcription factor, which may explain the pleiotropic but cell-type-specific action of the GR [6].

The dual role of GCs has been well documented [9]. In animal models, GCs protected against cancer development, and several studies point toward a tumor-suppressive role of GR in epithelial solid cancers [9, 177]. However, GR action in cancer biology appears cancer/cell-type dependent and influenced by treatment [64].

Although GCs and GR-mediated actions are intensively investigated, their pleiotropic, context-dependent, and cell-type-specific actions are not entirely understood. Their role in breast cancer, despite recent advances, is complex and still unpredictable. To clarify and specify this action of GC/GR system, the level of GR expression, the detection of its splice variants and posttranslational modifications, and its subcellular localization should be further investigated in clinical specimens.

Seemingly, the presence of ER is a key regulator of GR action, and their crosstalk was proposed to be responsible for the differential impact of GR expression and activity across breast cancer subtypes (ER + vs. TNBC). However, controversial preclinical and clinical findings will likely lead to further studies to understand the underlying causes. Also, the high context dependency of GC action represents a principal challenge in the discovery of further connections.

As in breast cancer, GCs are frequently used to mitigate undesirable side effects of conventional chemotherapy, and recent findings that GCs can promote tumor development and metastasis have raised understandable concerns.

Due to this pleiotropic action, as recently suggested, targeting GR activity is not a favorable therapeutic option [6]. This is further supported by the relevance of hierarchical network modeling in cancer drug-target selection [224]. In an interesting study applying hierarchical network structure to biological networks representing genome-scale dynamics, authors were able to faithfully model the therapy response by analyzing FDA-approved drugs compared to drugs that have been rejected [225]. If global transcriptional and protein–protein interactions are considered as a network, genes and proteins in such network represent nodes, and the interactions between them represent edges. Hubs are defined as the top 20% of the highly connected nodes. In a hierarchical network, genes/proteins in the top layer are the master regulators of the network because they influence the whole network through their downstream targets [226]. The core layer is the most abundant layer because it contains most of the HUBs and network motifs (interactions). The core layer plays a central role in the regulation of signal propagation. The bottom genes are located in the third layer, which directly regulates the genes of effector molecules, or they are the effectors of the network [225, 227]. Although nodes in the top and core layers are more likely to be important for the successful survival and adaptation of the organism to its environment, it has been demonstrated that drugs targeting the higher levels of a hierarchical network were more influential [224], while those targeting HUBs are not so effective [224]. In a hierarchical network, GR can be considered as a hub showing typical hub characteristics, such as regulating or co-regulating a high amount of all members of the network [225]. Indeed, GR is regulated by several factors and regulates a significant amount (10–20%) of the whole transcriptome, hence have numerous interactions in a hierarchical network [225]. Therefore, GR targeting therapy in a similar analysis may fall into the not efficient drug category; however, GR targeting in combination could be beneficial.

Therefore, in the use of GC and its agonists in breast cancer, the GR’s context-dependent functions have to be kept in mind. However, GR selective modulators and targeting GR co-regulators probably will be promising and effective therapeutical options, a conclusion that is supported by the current number of clinical trials investigating SEGRAMs or Hsp90 inhibitors.

With the growing body of evidence related to the potential correlation of GR with prognosis and response to therapy, as suggested earlier, the assessment of GR (and other steroid receptors) status in tumor tissue (quadruple and/or quintuple with AR) may add further possibilities for patient’ classification in order to select that best therapeutical approach, i.e., endocrine agonist/antagonist therapies and chemotherapy [3].