Looking beyond the ER, PR, and HER2: what’s new in the ARsenal for combating breast cancer?

Abstract

Breast cancer (BrCa) is a complex and heterogeneous disease with diverse molecular subtypes, leading to varied clinical outcomes and posing significant treatment challenges. The increasing global burden of BrCa, particularly in low- and middle-income countries, underscores the urgent need for more effective therapeutic strategies. The androgen receptor (AR), expressed in a substantial proportion of breast cancer cases, has emerged as a potential biomarker and therapeutic target. In breast cancer, AR exhibits diverse functions across subtypes, often interacting with other hormone receptors, thereby influencing tumor progression and treatment responses. This intricate interplay is further complicated by the presence of constitutively expressed AR splice variants (AR-Vs) that drive resistance to AR-targeting therapies through structural rearrangements in the domains and activation of aberrant signaling pathways. Although AR-targeting drugs, initially developed for prostate cancer (PCa), have shown promise in AR-positive breast cancer, significant gaps remain in understanding AR’s precise functions and therapeutic potential. The systemic management of breast cancer is guided primarily by theranostic biomarkers; ER, PR, HER2, and Ki67 which also dictate the breast cancer classification. The ubiquitous expression of AR in BrCa and the emergence of AR-Vs can assist the management of disease complementing the standard of care. This article provides a comprehensive overview of AR and its splice variants in the context of breast cancer, highlighting their prognostic and predictive value across different subtypes looking beyond the conventional ER, PR, and HER2 status. This review also raises the possibility of using AR splice variants in predicting tumor aggressiveness. From the settings of developing nations, this may provide useful insight by integrating recent advances in AR-targeted therapies and exploring their translational potential, emphasizing the critical need for further research to optimize AR-based therapeutic strategies for breast cancer management.

Background

The androgen receptor, also known as NR3C4, (nuclear receptor subfamily 3, group C, gene 4) belongs to the steroid receptor superfamily. Other members of steroid receptor groups are Estrogen Receptor (ER), Progesterone Receptor (PR), Glucocorticoid Receptor (GR), and Mineralocorticoid Receptor (MR) [1, 2]. Being a ligand-dependent transcription factor, AR mediates the biological effects of androgens, primarily testosterone and dihydrotestosterone (DHT). It plays a vital role in various physiological processes, including developing male reproductive systems, secondary sexual characteristics, and overall homeostasis. Despite being extensively studied in the context of prostate cancer, the role of AR in breast cancer has garnered increasing attention. AR having a substantial expression, complex, and multifaceted role in diverse molecular subtypes of breast cancer may influence tumor behavior, response to therapy, and clinical outcomes [3–5]. This complexity is aggravated by AR splice variants (AR-Vs) which contribute to the therapy resistance documented in prostate cancer. Given the pivotal role of AR in breast cancer, it is imperative to explore its potential as a prognostic and predictive biomarker, as well as a therapeutic target.

This review delves into the structural intricacies of the AR gene, highlighting the significance of its splice variants and their functional implications. Understanding androgen signaling, as well as the structural, functional, and clinical dimensions of AR and its variants, is essential for developing effective, tailored treatments and improving patient outcomes.

Circulating androgens in women

Androgens, primarily, are male sex hormones, which are also produced in smaller quantities in females and are crucial for regular reproductive health and fertility [6, 7]. They were first described in 1889, by French physiologist and Professor of Medicine Charles Edward Brown-Sequard from testicular extracts [8]. In women, circulating androgens are secreted by the adrenal glands as dehydroepiandrosterone-sulphate (DHEAS), dehydroepiandrosterone (DHEA), and androstenedione (A4), and by the ovaries as testosterone, DHEA, and A4 as outlined in Fig. 1 [9]. Peripheral tissues, such as the brain, bone, and breast, also produce testosterone and DHT [7]. Within cells, testosterone is converted to more potent androgen, DHT, by the enzyme 5α-reductase [10, 11]. The aromatase enzyme in various tissues, including mammary glands, converts testosterone to estradiol. DHT differs from testosterone by lacking a double bond in ring A [12], increasing its affinity for the AR two-fold and reducing the rate of dissociation of the AR-DHT complex five-fold compared to the AR-testosterone complex [13]. In women, circulating androgen levels decline during mid-reproductive years and continue to decrease after menopause [14]. Elevated testosterone levels are associated with an increased risk of breast cancer in pre-menopausal women.

Fig. 1.

Biosynthesis and interconversion of circulating androgens in women. Androgens, including DHEAS, DHEA, A4, and testosterone, are produced by the adrenal glands and ovaries. Testosterone and DHT (with a higher affinity for the AR) play a significant role in androgen signaling by binding to and catalyzing ligand-dependent action of AR. Created in https://BioRender.com

Structural components of AR

The AR gene, located on the X chromosome at locus Xq12, spans 186,599 base pairs and includes 8 exons, resulting in a 10,667 bp mRNA transcript encoding a 920-amino acid protein with a molecular weight of approximately 110 kDa [15, 16]. Like other steroid receptors, AR is characterized by three major functional domains.

The N-Terminal domain (NTD)

The NTD, encoded by exon 1, comprises about 60% of the AR protein, encompassing residues 1 to 555. Studies involving AR deletion mutations have shown that NTD is crucial for the function and transactivation of AR [17]. It contains the constitutively expressed Activation Function domain 1 (AF-1), which includes two separable transcription activation units: the ligand-dependent TAU-1 and ligand-independent TAU-5 [18, 19]. These subunits facilitate the transcriptional activity of AR and interaction between the NTD and the Ligand Binding Domain (LBD) [20]. This NTD-LBD interaction stabilizes the AR dimer, and ultimately, transcriptional regulation [21]. The NTD is linked to the pathogenesis of several diseases with increased mutations mapped in this region [22].

AR-NTD incorporates two polymorphic trinucleotide repeat segments, the first of which are highly variable poly-glutamine (poly-Q) repeats [23]. These poly-Q sequences regulate folding and structure within the NTD, thus affecting AR-transcriptional activity [24]. While a majority of studies have not linked poly-Q with BrCa risk, Rebbeck et al. have demonstrated that women carrying AR allele with ≥ 28 CAG repeats were at higher risk of breast cancer [25]. The NTD also harbors poly-glycine (poly-G) repeats. These, however, are researched in limited studies with no relevant correlations to breast cancer risk [26, 27].

DNA binding domain (DBD)

The DBD, comprising residues 556 to 623 and encoded by exons 2 and 3, is a cysteine-rich, highly conserved region among the steroid receptor superfamily. The crystal structure of the DBD shows two zinc fingers, each containing four cysteine residues responsible for recognizing and binding to promoter DNA sequences called the Androgen Response Element (ARE), thereby regulating gene expression. The α-helix of the zinc finger interacts with the nucleotides in the DNA major groove of ARE [28, 29].

The DBD is connected to the LBD by a flexible hinge region (residues 623–665), which includes a Nuclear Localization Signal (NLS) facilitating AR nuclear import upon ligand binding [30]. AR, being a large protein of 110 kDa, requires active transport into the nucleus upon ligand binding. Li et al. suggested that androgen binding induces a switch exposing the NLS which ultimately facilitates nuclear entry of AR by importin family of proteins [31]. The NLS contains two bundles of basic amino acids that are highly conserved in the steroid receptors superfamily [21]. One motif in the hinge region, 629-RKLKKL-634, plays a role in DNA binding, AR translocation, and coactivator recruitment [32].

Ligand binding domain (LBD)

The C-terminal LBD, encompassing residues 666 to 920 and encoded by exons 4 to 8, is crucial for recognizing and binding androgens such as testosterone and DHT. Eleven α-helices and four short β-strands making two anti-parallel β-sheets constitute the AR LBD [21]. The Activation Function domain 2 (AF-2) is located on a ligand binding pocket, which functions as the binding site for endogenous androgens. The AF-2 facilitates cross-talk between AR domains by providing interaction between the N- and C-termini of the receptor [18]. When androgens bind to the AR-LBD, the AF-2 region becomes crucial for initiating a cascade of molecular events that leads to the recruitment of coactivators, binding thereof, and transcriptional activation of androgen-responsive genes.

The LBD also interacts with various agonists and antagonists, making it a significant target for therapeutic interventions in hormone-responsive cancers, such as prostate cancer. Mutations in the LBD can alter androgen binding, receptor activation, and gene expression, potentially contributing to the development and progression of prostate cancer [33]. Due to its key role in AR signaling, the LBD is a prime target for therapeutic interventions.

AR signaling

Being a steroid hormone-activated transcription factor, AR is dependent on androgens for its activity. It, however, can catalyze its actions either by DNA binding or independently. AR is located in the cytoplasm sequestered with the Heat Shock Proteins (HSP) and other chaperons [34]. Circulating androgens (testosterone), being steroid hormones, may passively diffuse through the cell membrane. Testosterone may get converted into its active metabolite, DHT by 5α-reductase. The binding of testosterone or DHT to the ligand-binding pocket promotes conformational changes in the AR that expose the NLS.

As exhibited in Fig. 2, through DNA binding-dependent action, commonly known as classical AR signaling cascade, AR disassociates from HSP and accessory proteins. Following this, AR translocates into the nucleus, dimerizes [35], and then attaches to the androgen response element in the promoter region of target genes, causing modulations in DNA transcription. Recruiting other proteins and coactivators, AR controls gene transcription positively or negatively, leading to a cascade of molecular events such as metabolism, differentiation, proliferation, apoptosis, or angiogenesis [36, 37].

Fig. 2.

Classical AR signaling in an androgen-responsive cell. AR activation through its ligands regulates gene expression by AR binding to androgen response elements on DNA, thus initiating the transcription of AR-responsive genes. Created in https://BioRender.com

However, in addition to its DNA binding-dependent action, AR also exhibits DNA binding independent activities that play a crucial role in cellular function. In the DNA binding independent events, AR can initiate rapid, non-genomic signaling by activating second messenger pathways such as MAPK (Mitogen-Activated Protein Kinase) pathway and PI3K (Phosphoinositide 3-Kinase)/Akt pathway [38].

Clinical relevance of AR in cancer settings

The AR is necessary for the proper growth and functioning of the prostate and plays a crucial role in the development and progression of prostate cancer, one of the most common cancers in men. In their seminal work, Huggins and Hodges correlated the AR-dependent growth and survival in PCa [39]. AR expression has ubiquitously been linked to both primary and metastatic PCa, making it a key target in treatment [40]. Androgen Deprivation Therapy (ADT) is the cornerstone of advanced PCa treatment [41]. ADT approaches include using drugs that inhibit androgen production, block their interaction with AR, or through surgical removal of the testes to eliminate the primary source of androgen production [42]. Bicalutamide, an AR antagonist, competes with endogenous androgens at the LBD, inhibiting AR activation [43]. Enzalutamide, a second-generation non-steroidal AR antagonist [44], has a higher affinity for AR and inhibits nuclear translocation of AR as well [45]. It has shown a significantly low risk of relapse in a double-blind randomized phase 3 trial [46]. With different modes of action focussing on AR signaling, AR antagonists are essential components of the treatment of prostate cancer and are useful in the management of this disease.

However, the efficacy of ADT can be limited by the development of resistance over time, leading to an advanced form of PCa, known as Castration-Resistant Prostate Cancer (CRPC) [47]. CRPC remains androgen-driven through multiple mechanisms such as AR truncations, mutations, overexpression, amplification, and extragonadal androgen synthesis, including within cancer tissue itself [48]. Several reports have linked constitutively expressed androgen receptor splice variants (AR-Vs) as a potential cause of treatment resistance in CRPC [49–53].

Alternative splicing in cancer

Alternative splicing (AS) is a fundamental molecular process joining exons in various combinations, prevalent in over 90% of human genes [54]. This process creates multiple mRNA isoforms from a single gene, contributing to genetic diversity [55–58]. Splicing mechanisms are orchestrated mainly by exon skipping, mutually exclusive exons, alternative splice sites, and intron retention as illustrated in Fig. 3 [59–62].

Fig. 3.

Commonly observed alternative splicing patterns. Differential combinations of exonic and intronic sequences generate full-length transcript or splice variants. Exons and introns are represented by colored boxes and horizontal lines, respectively while splicing arrangement is represented by dotted lines

Dysregulation of splicing factors commonly exhibited by cancer cells changes the splicing patterns, contributing to tumorigenesis [63–65]. AS can produce oncogenic isoforms that promote cell proliferation, inhibit apoptosis, and enhance metastasis [66]. Dysregulation of normal proliferation and apoptotic pathways takes place by aberrant splicing patterns in proteins such as the TP53 gene, which encodes the p53 tumor suppressor protein [67]. Certain splice variants of the BCL-X gene, a member of the BCL-2 family that regulates apoptosis, are involved in the evasion of apoptosis, metastasis, and resistance against chemo- or targeted therapies [68]. The AS of genes such as CD44 and VEGF-A have been reported to aggravate Epithelial-Mesenchymal Transition and angiogenesis, respectively [69]. Splice variants in several crucial genes can modify signaling pathways that regulate cell cycle, growth, differentiation, and survival facilitating tumor progression [70].

AR splice variants

Among AR splice variants, AR-Vs, particularly AR variant 7 (AR-V7), are extensively studied for their clinical relevance as prognostic and potential predictive biomarkers. AR-Vs lacking the LBD contribute to drug resistance in ADT, as anti-androgen drugs target the LBD. The earliest report of AR-Vs comes from Gregory et al., who in 2001 provided the first evidence of AR-Vs’ endogenic expression in PCa cell lines. They observed shorter molecular weight bands in the 80–90 kDa range on AR immunoblots, which were more prominent in CRPC cell lines [71]. AR45, an NTD truncated isoform, was identified in 2005, predominantly in the heart and other tissues, including the prostate, breast, and lung [72]. In the late 2000s, AR splice variants gained prominence, especially after the discovery of several variants lacking the LBD (except for AR-45). These variants can potentially facilitate constant AR activity. Identifying these variants laid the groundwork for subsequent studies crucial in comprehending CRPC’s molecular mechanisms [73, 74].

Hu et al. in their landmark work, identified cryptic exons (CE1, CE2, CE3) in intron 3, upstream to the DBD of AR. Alternative splicing involving these cryptic exons leads to seven novel AR-Vs (AR-V1 to AR-V7), which lack the DBD. AR-V1 and AR-V7 are abundantly expressed in clinical prostate cancer cases, with approximately 20-fold higher levels observed in CRPC [75]. Watson et al. in 2010, using conventional and next-generation sequencing approaches not only validated the known AR-V1 and AR-V7 but also detected several structurally diverse AR isoforms. They also reported the upregulation of AR-V1 and AR-V7 after castration at both mRNA and protein levels [76]. The discovery of AR-V1 and AR-V7 was confirmed by Guo et al. and further established by immunoblot in different cell lines [77]. AR-Vs validated and annotated on the NCBI database (as of November 2024) are summarized in Fig. 4 [15].

Fig. 4.

Organization of the AR gene and its splice variants annotated on NCBI. Located on the X chromosome, the AR primary transcript translates into AR-FL and various AR splice variants. Except for AR-TV2, most of the AR-Vs have ligand binding domain truncated. This figure presents the structural complexity of the AR, highlighting its relevance in cancer biology and potential therapeutic targeting. Created in https://BioRender.com

Depending on their conditional expression, most AR-Vs can be categorized into three groups: (1) constitutively expressed, (2) conditionally expressed, and (3) inactive [73]. The most extensively studied variant is AR-V7 in terms of clinical relevance. Since its discovery, AR-V7 has garnered more attention due to its constitutively high expression in CRPC cases [73, 78], and nuclear transport [79]. Additionally, AR-V7 expression significantly goes up following AR-FL signaling suppression due to androgen depletion, AR-FL knockdown, or enzalutamide treatment in VCaP and LNCaP95 cells but not in LNCaP and CWR22Rv1 cells. This indicates that the control of AR variant levels may rely on the specific cellular context alongside the cell-specific roles of AR splice variants [80]. Some AR-Vs exhibit conditional transcriptional activity, dependent on specific cellular contexts. For example, AR-V1 and AR-V9 are active in AR-FL-expressing cells (LNCaP) but not in AR-FL-negative cells (PC3) [81]. Inactive AR-Vs such as AR-V13 and AR-V14 are reported to have a partially truncated AR-LBD [81].

Mode of action of AR-Vs

AR-Vs exhibit varied nuclear transport capabilities due to the loss of the NLS. Immunofluorescent staining evidences the constitutive nuclear localization of AR-V7 even in the absence of androgens, while AR-V1, AR-V9, and AR-V13 remain cytoplasmic [75, 76, 81]. To explore the nature of AR-V7 nuclear transport and its transcriptional function, Chan et al. demonstrated a resemblance between unique regions in the AR-V7 C-terminal and AR-NLS. Mutations in some residues of this unique region compromised the AR-V7 nuclear transport [82].

Homodimerization is essential for the optimal transcriptional activity of AR-FL, involving intramolecular N-terminal and C-terminal interactions leading to nuclear transport and downstream actions [83]. Similar to AR-FL, AR-Vs require dimerization for transactivation of target genes, which can occur through homodimerization or heterodimerization with other AR-Vs [84]. AR-FL and AR-V7 heterodimerization can facilitate AR-FL nuclear localization and transcriptional activity regardless of androgens [52]. However, there is no consensus on the binding sites of AR-FL homodimer and AR-FL/AR-V7 heterodimer, with some studies showing overlapping regions and others indicating unique bindings [85–88]. AR-V7 binds to unique transcription sites through interactions with co-regulators like HOXV13 and ZFX [87, 88]. The absence of the LBD in AR-V7 is likely to substantially alter its 3D structure in AR-FL/AR-V7 heterodimers, exposing or masking regions that interact with co-regulators. Thus, while AR-FL and AR-V7 share some targets, AR-V7 may regulate distinct gene sets [75, 77, 80, 87, 89].

Clinical implications of AR-Vs

Prognostic and predictive value of AR-Vs

AR-Vs, especially AR-V7, have emerged as valuable prognostic markers for predicting resistance to AR-targeted therapies like abiraterone and enzalutamide. AR-V7 mRNA expression in circulating tumor cells (CTCs) predicts outcomes in CRPC patients, with AR-V7-positive CTCs showing resistance to next-generation AR-directed agents and less favorable clinical outcomes [51, 90]. Elevated levels of AR-V7 in hormone-naïve PCa correlate with an increased risk of relapse after radical prostatectomy [75, 77].

Prior studies in CTCs have examined how AR-V7-positive patients responded to taxanes versus enzalutamide or abiraterone. Taxanes were more effective in AR-V7-positive patients, especially in terms of Prostate-specific antigen (PSA) - Progression-Free Survival (PFS) and PFS, affirming CTC AR-V7’s predictive potential [91]. However, the prognostic value of AR-Vs in tissues is less clear compared to liquid biopsy, Zhu et al. have observed increased expression of AR-V7 and AR-V7/AR-FL ratio as the PCa progresses to CRPC [92, 93]. These results reinforce AR-V7 as a prognostic marker linked to poor outcomes in metastatic CRPC patients treated with abiraterone or enzalutamide [94].

Therapeutic potential of AR-Vs

Given AR-Vs’ significance in CRPC progression, there is a growing need for AR-Vs targeting therapies. While taxanes are currently the primary treatment choice for AR-V7-positive CRPC patients, research is underway to develop new drugs that target AR-Vs [95]. Such approaches aim to disrupt AR-Vs signaling, either by targeting AR-NTD [96], or inhibition of protein expression [97], reducing AR-Vs expression [98] or prevention at DNA-binding [99].

Ji et al. discuss NTD-targeting agents under development with a broad spectrum of effects on AR-FL and other AR-Vs containing AR-NTD [100]. For instance, EPI, non-steroidal AR antagonists that directly target the NTD of AR (which is also present in other AR-Vs) and prevent coactivators from binding to it, are under investigation [101, 102]. EPI-7386, one such novel drug is currently in clinical trials to evaluate its efficacy, both as a monotherapy and in combination with enzalutamide, for AR-V7-positive CRPC patients [95]. Additionally, niclosamide, an FDA-approved anti-helminthic drug, was found to selectively inhibit AR-V7 protein expression through enhanced protein degradation in preclinical studies [103]. Although not all of these compounds will likely find their way into clinics, these new medications offered novel and promising ways to efficiently target AR-Vs [104].

The study of AR splice variants in CRPC is still in the early stages despite significant interest in the years since their discovery. AR-Vs, notably AR-V7, are consistently detectable in various biospecimens and have negative prognostic implications. Although further clinical investigations with multiple treatment arms are needed to confirm the predictive value of AR-Vs, some results indicate that CTC-based AR-V7 may be a useful predictive biomarker when choosing between AR-targeted therapies and taxanes in treatment decisions [91, 105, 106].

AR in breast cancer

The heterogeneous nature of breast cancer poses significant treatment challenges, necessitating new therapeutic regimens in addition to currently employed drugs. Keeping aside methodological variability such as AR localization, IHC cut-off, and staining protocol used, AR is reported to be expressed in 70 − 90% of clinical breast cancer [3–5]. This ubiquitous expression pattern positions AR as an emerging marker and therapeutic target. Immunohistochemistry of AR in the prostate and different subtypes of breast cancer tissues is illustrated in Fig. 5.

Fig. 5.

Immunohistochemical staining of AR in the prostate and different subtypes of breast cancer at different magnifications. The varying intensity of AR expression across different subtypes highlights the receptor’s differential expression patterns and potential diagnostic significance. IHC performed with anti-AR mAb (MA5-16412, Invitrogen 1:400)

AR expression and relevance in different BrCa subtypes

Although there are similarities between prostate and breast cancers due to the hormone-driven nature of both diseases, there are considerable differences in AR signaling pathways and their downstream effects [107]. In PCa, AR functions as a primary driver of tumor growth, with AR-Vs such as AR-V7 contributing to therapeutic resistance. In contrast, AR’s role in BrCa varies by subtype, acting as a tumor suppressor in ER-positive subtypes, and a potential oncogene in AR-positive Triple-Negative Breast Cancer (TNBC) [108, 109]. These nuanced roles underscore the need to explore the subtype-specific implications of AR in BrCa while leveraging insights from PCa to guide therapeutic development.

Luminal type

A comprehensive systematic review of 19 studies involving 7,693 women found that AR co-expressed with ER-positive breast cancer in 74.8% of cases versus 31.8% in ER-negative tumors [110]. In further analyses based on molecular subtypes, Luminal types, characterized by ER-positivity, displayed varying levels of AR expression, ranging from 50 to 90% [111]. In breast cancer, there is a complex interplay between AR and ER. Normal breast tissue expresses two types of ER involved in reproductive organ development and other cellular processes. Among the two ERs, ERα and ERβ, with estradiol being their natural ligand, ERβ is dominant in normal breast. However, in breast cancer, ERα expression increases, contributing to tumorigenesis [112].

Preclinical studies indicate varying proliferative effects linked to the ratios of ERα and AR and the availability of their ligands [113–116]. Importantly, testosterone can also be peripherally converted to estradiol via the aromatase enzyme, adding complexity to the AR-ER relationship [7, 117, 118]. Panet-Raymond et al. have demonstrated an interaction between AR and ERα that might alter both receptors’ transcriptional activity, with AR having a greater impact on ERα transactivation [114]. Studies on AR activity in breast cancer carcinogenesis, primarily conducted in vitro using breast cancer cell lines like MCF-7, T47-D, and BT20, have demonstrated that AR antagonism has an antiproliferative effect [119–121]. This highlights the intricate and varying responses to hormones and their antagonists in breast cancer cells which could depend on the several proteins that interact with AR [122].

HER2 enriched

AR expression in HER2-enriched subtypes ranges from 20 to 60% [3, 111, 123]. Preclinical studies suggest that the molecular apocrine subtype (ER-negative, AR-positive) may proliferate in the presence of androgens, driven by intricate interactions between the AR and the HER2 signal transduction pathway, independent of the ER pathway [124]. HER2 kinase signaling is essential for full AR activity at low androgen levels, enhancing AR’s DNA binding and stability [125]. HER2-enriched and AR-positive cases correlate with poor clinical outcomes, including lower DFS and OS [126]. Another study finds an association between high AR mRNA with shorter DFS and OS in HER2-positive ER-negative patients [36]. DHT treatment in AR-positive HER2-enriched cell lines like MDA-MB-453 promotes growth through oncogenic Wnt and HER3 activation [127].

Additional research in cell lines and tissue samples emphasized the cross-regulation of specific genes between AR and HER2. This led to increased expression of steroid response genes and enhanced cell proliferation when either AR or HER2 was stimulated. Notably, the anti-androgen flutamide and HER2 inhibition demonstrated pro-apoptotic effects with a synergistic impact when used in combination [128]. The findings from these studies revealed that AR contributes to the carcinogenesis of HER2-enriched breast cancer.

TNBC

TNBC subtype displays comparatively lower AR expression (20–50%) [111, 123]. Traditionally, TNBC has exhibited poor outcomes, with a median overall survival of around 13 months in metastatic cases and a shorter time from disease recurrence to death compared to other breast cancers [129, 130]. TNBC remains biologically diverse despite its common features, complicating therapeutic target identification [131]. The Luminal Androgen Receptor (LAR) subtype, characterized by enriched AR signaling, accounts for 22% of TNBC cases [132, 133], and demonstrated increased sensitivity to the AR antagonist bicalutamide in in vitro studies [134]. AR-positive TNBC patients have lower survival rates than AR-negative ones [135]. Notably, even non-LAR TNBC cell lines showed involvement in AR signaling. AR was observed to up-regulate the EGFR ligand amphiregulin, promoting cell proliferation through the EGFR pathway, and the anti-androgen enzalutamide blocked this effect [136]. Thus, AR consistently promotes cellular proliferation in non-luminal types, making it a potential therapeutic target.

Prognostic and predictive value of AR in BrCa

Patients with ERα-positive and AR-positive tumors exhibit better outcomes due to AR competing with ERα for Estrogen Response Elements (ERE) [137], impairing ERα-induced transcription and promoting apoptosis [126, 135]. Clinical studies validate these findings, indicating that ERα-positive AR-positive patients have improved outcomes [126, 135, 138]. In contrast, in ERα-negative tumors, AR binds to androgen-responsive elements, promoting cell proliferation and tumor growth [138]. Ravaioli et al. report in ductal carcinoma in situ (DCIS), AR expression is higher in relapsed tumors, while ERα is higher in non-relapsed cases [139]. Thus, AR, along with hormone receptors (HR) may serve as a better prognostic marker to predict relapse in DCIS patients [140, 141].

The role of AR in luminal breast cancers is debated, with some studies identifying it as an independent prognostic biomarker [110, 142]. The AR/ERα ratio serves as a prognostic tool for in situ relapse or progression to invasive subtypes [140]. It can predict endocrine therapy response, with high nuclear AR/ERα ratios linked to hormone therapy failure, and enzalutamide effectively reduces tumor growth in hormone therapy-resistant cases [143]. However, in advanced breast cancers, AR is not a strong predictor of first-line anti-estrogen therapy effectiveness, with progesterone receptor and Ki-67 being more reliable predictors [111]. In ERα-positive, HER2-positive breast cancers, AR-positivity correlates with smaller, less aggressive tumors [144]. AR promotes growth in ERα-negative and HER2-positive cancers by affecting HER3 gene expression [127].

Clinical data suggested AR might not be a suitable biomarker for endocrine therapy selection. On the other hand, studies supported AR’s predictive value for AR inhibitors, especially in TNBC [145–147]. In TNBC, AR-positivity is associated with better disease-free and overall survival, particularly in the LAR TNBC subgroup [148]. Antiandrogen therapy may benefit AR-positive TNBC patients, whereas AR-negative, EGFR-positive TNBCs are high-risk and benefit from chemotherapy [149]. ERβ is expressed in approximately 30% of TNBC and can downregulate AR expression [150].

Treatment options

AR activation inhibits cellular proliferation in most ER-positive breast cancers but stimulates growth in ER-negative cases. This suggests a potential use of AR agonists or antagonists in treatment. In ER-positive and some ER-negative BrCa cases where AR activation prevents tumor growth, androgens have been employed for therapy [151], but they come with side effects [152]. Selective AR modulators (SARMs) like non-steroidal anabolic agent, enobosarm (GTx-024), with minimal side effects, have gained attention for ER-positive advanced BrCa treatment [153]. SARMs were shown to reduce tumor load by 90% in in vivo studies conducted in MDA-MB-453 cells [154].

AR antagonists, such as bicalutamide and enzalutamide, are commonly used for advanced breast cancer, especially in tamoxifen-resistant and TNBC cases, with positive clinical trial results [155–157]. For TNBC and HER2-enriched BrCa, AR can promote growth when combined with other factors involving cell cycle progression and intracellular pathways for survival, proliferation, and invasiveness. Combining AR antagonists with pathway inhibitors may be a promising approach to achieve optimal outcomes [158–160]. There are several clinical trials (listed in Table 1) modulating AR activity using AR inhibitors as monotherapy or combined with other medications in breast cancer settings.

Table 1.

Ongoing clinical trials targeting AR-positive breast cancer

Study identifier	Registration Date	Title	Breast cancer subtype	Trial agent	Agent class	Phase	Enrolment (n)	Primary outcome
NCT02605486	2015-11-11	Palbociclib in Combination with Bicalutamide for the Treatment of AR(+) Metastatic Breast Cancer	AR + BrCa	Palbociclib + Bicalutamide	CDK4/6 inhibitor	I/II	46	PFS
NCT02750358	2016-05-19	Feasibility Study of Adjuvant Enzalutamide for the Treatment of Early Stage AR (+) Triple Negative Breast Cancer	AR + TNBC	Enzalutamide	AR inhibitor	II	50	TDR
NCT02955394	2017-09-21	Preoperative Fulvestrant with or Without Enzalutamide in ER+/​Her2- Breast Cancer	AR+/ER+/Her2- BrCa	Fulvestrant + Enzalutamide	SERD	II	61	PEPI score
NCT03090165	2018-05-07	Ribociclib and Bicalutamide in AR + TNBC	AR + TNBC	Ribociclib + Bicalutamide	CDK4/6 inhibitor	I/II	37	MTD, CBR
NCT03650894	2019-04-03	Nivolumab, Ipilimumab, and Bicalutamide in Human Epidermal Growth Factor (HER) 2 Negative Breast Cancer Patients	HER2-	Nivolumab + Bicalutamide and Ipilimumab	anti-PD-1 + AR inhibitor and anit-CTLA4	II	30	CBR
NCT04360941	2020-08-11	PAveMenT: Palbociclib and Avelumab in Metastatic AR + Triple Negative Breast Cancer (PAveMenT)	AR + TNBC	Palbociclib + Avelumab	CDK4/6 inhibitor and immunotherapy drug	I	45	MTD, ORR
NCT04947189	2022-11-01	Seviteronel in Combination with Chemotherapy in Androgen-receptor Positive Metastatic Triple-negative Breast Cancer (4CAST)	AR + TNBC	Seviteronel + Dexamethasone and Docetaxel	AR inhibitor	I and II	65	DLT
NCT05095207	2021-09-20	Abemaciclib in Combination with Bicalutamide for Androgen Receptor-positive, HER2-negative Metastatic Breast Cancer	AR + HER2 -	Abemaciclib + Bicalutamide	CDK4/6 inhibitor	IB/II	60	DLT, RP2D
NCT05573126	2023-01-11	Study to Evaluate EP0062 in Patients with Relapsed Locally Advanced or Metastatic Androgen Receptor Positive (AR+)/​HER2-/​ER + Breast Cancer	AR+/HER2-/ER + BrCa	EP0062	SARM	I/II	128	DLT, MTD, AE, and SAE
NCT05673694	2023-03-08	To Evaluate a Phase Ia/​Ib Clinical Study of EG017 in Patients with Advanced Breast Cancer	AR+, ER+, HER2-	EG017	SARM	Ia/Ib	70	DLTs, AE, and SAE
NCT05954442	2023-09-13	Everolimus With Investigator’s Choice of Chemotherapy in Advanced Triple-Negative Breast Cancer (TNBC) With LAR Subtype (BCTOP-T-M03)	AR + TNBC	Everolimus + Investigator’s Choice of Chemotherapy	mTOR inhibitor	III	203	PFS
NCT06099769	2023-10-18	A Study of Enzalutamide, Enzalutamide in Combination with Mifepristone, or Chemotherapy in People with Metastatic Breast Cancer	AR + TNBC or ER-low BrCa	Mifepristone ± Enzalutamide ± Chemotherapy	GR and AR inhibitors	II	201	PFS
NCT06365788	2024-04-08	Bicalutamide and Abemaciclib in Inoperable or Metastatic Androgen Receptor-positive Triple-negative Breast Cancer (ABBICAR)	AR + TNBC	Abemaciclib + Bicalutamide	CDK4/6 inhibitor	II	53	DCR

Abbreviations: PFS Progression-free survival, TDR Treatment discontinuation rate, PEPI Preoperative Endocrine Prognostic Index, CBR Clinical Benefit Rate, MTD Maximum Tolerated Dose, ORR Objective Response Rate, DLT Dose Limiting Toxicity, RP2D Recommended Phase II Dose, AE Adverse Events, SAE Serious Adverse Events, DCR Disease Control Rate

AR-Vs in breast cancer

With the significant prevalence of AR in breast cancer several AR-targeted therapies are employed at the clinical stage. Despite exhibiting encouraging outcomes, it leads to medication resistance in certain patients, eventually worsening the disease. Although the mechanisms driving treatment resistance are still being investigated, there is a chance that they are similar to those seen in prostate cancer. One such mechanism is attributed to the generation and overexpression of AR splice variants [76]. While AR-Vs testing has been well-established in the clinical management of PCa, its application in BrCa remains in its infancy. Pioneering works by Hu and Hickey et al. demonstrated several AR-Vs in different classes of breast cancer and cell lines [161, 162]. AR-V7, the most prominently expressed variant, is linked to resistance to ADT, correlates with the HER2-enriched subtype, and is supported in PCa cells as well [125, 127, 128]. Another study conducted by Ferguson et al. reviewed the prevalence and clinicopathologic features associated with AR-V7 in a sizable BrCa cohort. Their report reassures AR-V7 testing for AR-positive cases in both primary and metastatic settings. This study further encourages the consideration of AR-V7 as a predictive biomarker for the benefit of AR antagonists [163]. Recently we have reported the association of AR-V7 with aggressive clinicopathological parameters and poor overall and disease-free survival in breast cancer patients [164]. Armstrong et al. discussed the strategies targeting AR-Vs, the efficacy of which is under evaluation [165]. Such therapeutics can be extended to AR-Vs positive BrCa settings.

The clinical utility of AR-V7 testing in different subtypes of BrCa is limited by the lack of attention to the clinical utility of AR-V7 and robust validation studies. The detection of AR-V7 in CTCs through Immunohistochemistry (IHC) or qRT-PCR as a tool for non-invasive biomarker development has been unexplored in breast cancer [166]. Current research relies predominantly on preclinical works, cell lines, and small patient cohorts with limited diversity, which restricts the generalizability of findings. Moreover, there is limited evidence on whether AR-V7 expression correlates consistently with therapeutic resistance in BrCa, as observed in PCa. Despite these challenges, the advancement of the EPI series of drugs that target AR-NTD has paved the way for targeted therapies with AR-V7-positive BrCa [167]. Moving forward, larger, well-designed clinical studies are needed to establish AR-V7 as a reliable biomarker for therapeutic decision-making in BrCa. Understanding the mechanisms and signaling pathways associated with these variants is crucial for developing targeted therapies that can effectively combat hormone-resistant breast cancer.

Conclusion

The AR is a significant player in breast cancer pathophysiology, with its diverse roles and interactive landscape across the different molecular subtypes. The advent of AR-Vs in mediating therapeutic resistance necessitates further exploration in the context of breast cancer. The development of SARMs and AR-NTD antagonist drugs has opened new avenues for targeted therapy with the potential for improved efficacy and fewer side effects. Future research should focus more on developing predictive biomarkers to stratify patient selection for anti-AR therapies, exploring AR signaling pathways in various breast cancer subtypes, and the interplay between AR and other hormone receptors. Additionally, understanding the functional significance of AR-Vs in breast cancer progression and treatment resistance will be essential to optimize therapeutic strategies. These efforts hold the promise of advancing precision medicine approaches, improving prognostic and predictive accuracy, and ultimately enhancing outcomes for breast cancer patients.

Abbreviations

ADT

Androgen Deprivation Therapy

AF

Activation Function

AR

Androgen Receptor

ARE

Androgen Response Element

AS

Alternative Splicing

bp

Base Pairs

CE

Cryptic Exon

CRPC

Castration-Resistant Prostate Cancer

CTC

Circulating Tumor Cells

DBD

DNA Binding Domain

DCIS

Ductal Carcinoma in situ

DFS

Disease-Free Survival

DHEA

Dehydroepiandrosterone

DHEAS

Dehydroepiandrosterone Sulphate

DHT

Dihydrotestosterone

EGFR

Epidermal Growth Factor Receptor

ER

Estrogen Receptor

HER2

Human Epidermal Growth Factor Receptor 2

IHC

Immunohistochemistry

LAR

Luminal Androgen Receptor

LBD

Ligand Binding Domain

NLS

Nuclear Localization Signal

NTD

N-Terminal Domain

OS

Overall Survival

PCa

Prostate Cancer

PFS

Progression-Free Survival

PR

Progesterone Receptor

PSA

Prostate-Specific Antigen

qRT-PCR

Quantitative Reverse Transcription Polymerase Chain Reaction

SARM

Selective Androgen Receptor Modulator

SERD

Selective Estrogen Receptor Degrader

TNBC

Triple-Negative Breast Cancer

Authors' contributions

TPS conceptualized the idea, created illustrations, and wrote the primary manuscript and drafts. RD and SK conceptualized, monitored, and oversaw the whole study. All authors read and approved the final manuscript.

Funding

The preparation of this review article required no funding.

Data availability

No datasets were generated or analysed during the current study.

Declarations

Ethics approval and consent to participate

No ethics approval and consent to participate were required for this review article, as it does not involve primary data collection involving human participants or animals.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.