Explaining hormone-deprivation therapy failure in prostate cancer patients

In 1941, Huggins and Hodges demonstrated the clinical efficacy of hormonal manipulation for the treatment of metastatic prostate cancer [1]. Hormone deprivation therapies employing surgical and/or medical castration, as well as their combination with anti-androgens, have since become the mainstay of systemic treatment for advanced prostate cancer. In a contemporary clinical setting, the length of clinical remission, often assessed by serum prostate-specific antigen (PSA) measurements, varies substantially due to a wide-spectrum of clinical phenotypes among treated patients [2]. Almost invariably, however, prostate cancer develops castration-resistant phenotype and progresses to a life-threatening stage, despite hormone therapies. The widespread use of hormone deprivation therapies is manifested in the observation that almost all patients who died from prostate cancer had received and failed hormone-deprivation therapies. Understanding the molecular underpinnings that determine the castration-resistant phenotype is, therefore, of utmost importance in developing new clinical tools to predict, monitor or control therapeutic resistance and to treat relapsed, castration-resistant prostate tumors.

The androgenic-signaling pathway has been the focus of attention in efforts to delineate the biologic and molecular determinants of castration resistance. Androgenic functions are mediated through the androgen receptor (AR), a member of the steroid hormone receptor transcription factors. Since the cloning of the human AR approximately 20 years ago, much has been learned about the complex molecular components of androgenic signaling [3]. The C-terminal ligand-binding domain of the AR protein binds androgens, such as testosterone and dihydrotestosterone, inducing AR conformational changes that allow its entry into the nuclei. In the cell nucleus, AR binds to DNA sequences known as androgen response elements, initiating and facilitating transcription of genes that are the ultimate functional effectors of androgenic signaling essential for prostate cell growth and survival. A few lines of compelling evidence have established that, unlike human breast cancer, prostate cancer progression upon hormone therapy is not due to loss of dependence on hormonal signaling but, instead, characterized by sustained androgenic signaling that bypasses the requirement for physiological levels of androgens. First, with rare exceptions, prostate cancer patients dying from castration-resistant prostate cancer have very high levels of serum PSA [4], the production of which is driven by androgenic signaling. Second, castration-resistant prostate cancers have elevated expression levels of the key mediator of androgenic signaling, the AR, and this is a very consistent molecular feature in tissues derived from patients with castration-resistant prostate cancer [5]. Third, a subset of prostate cancers that relapsed following hormone therapy continue to respond to second-line hormone therapies designed to disrupt the AR signaling axis, suggesting that AR-mediated androgenic signaling is still operating among these tumors [6]. While it is possible that AR-negative prostate cancer cells may give rise to androgen-independent prostate carcinoma [7], prostate tumors comprised of mainly AR-negative malignant cells (i.e., small cells and neuroendocrine cells) are extremely rare. Therefore, the question “why do prostate cancer patients fail hormone therapy?” can be paraphrased as “why does AR signaling persist despite hormone therapy?”

Multiple mechanisms leading to persistent AR signaling in castration-resistant prostate cancer have been proposed. These include AR amplification and overexpression [8], somatic AR mutations allowing AR to be activated upon nonspecific, promiscuous ligand binding [9], altered AR partner molecules, such as AR coactivators and corepressors [10], or AR activation through crosstalk with other signaling pathways that are generally enhanced in cancer cells [11]. Importantly, all these mechanisms operate in the presence of AR ligands. With regard to the requirement for AR ligands, existing hormone therapy regimens do not lead to complete ablation of androgens. In addition, controversial yet interesting new data have emerged that suggest that prostate cancer cells can make their own androgens, which may mediate intracrine androgenic signaling [12]. Thus, the overall prevailing explanation for hormone therapy failure is that the AR signaling axis may be only temporarily and partially impaired during the acute phase of hormone ablation but later restored following the re-emergence of castration-resistant prostate cancer, owing to continued production and supply of androgens, albeit at lower levels and altered types, in combination with elevated expression and function of multiple molecular components in the AR signaling pathway. Based on this prevailing ligand-dependent mechanism, a number of therapeutic approaches have been developed and are under clinical trials to treat castration-resistant prostate cancer [5].

These new therapeutic approaches under development are largely based on the assumption that AR activation depends on the presence of AR ligands and also requires an intact AR ligand-binding domain. A few very recent studies added a new twist on the possible mechanisms leading to AR activation in castration-resistant prostate cancer. These studies established that AR signaling in a clinically relevant context could occur in the complete absence of ligand binding [13,14], due to the presence of novel AR variants specifically produced in castration-resistant prostate cancer cells. These novel AR variants are encoded by spliced transcripts and do not have the protein domains needed to bind to androgens but are constitutively active and drive AR signaling in the complete absence of androgens. Studies in my laboratory have focused on AR-V7, one of the variants with the most abundant expression and also the highest activity [13]. AR-V7 was elevated by approximately 20-fold in castration-resistant prostate cancer cells derived from patients who died from metastatic prostate cancer following hormone therapy failure. More interestingly, generally lower but varied AR-V7 expression was also detected in prostate cancers that had not been influenced by hormone ablation, and higher AR-V7 expression predicted PSA recurrence following local therapy in these patients [13]. These results suggest that castration-resistant prostate cancer cells bearing the signatory marker of a constitutively active AR are present prior to hormone therapies, and these cells may propagate under the selection pressure induced by lack of sufficient androgens, leading to progressive castration-resistant prostate cancer.

Following decoding of the variant-specific peptide sequences and development of variant-specific antibodies against AR-V7, both detection and manipulation of this castration-resistant AR variant are made possible [13]. Detection and manipulation of such molecular indicators of castration-resistant prostate cancer cell will have many important clinical implications. First, an established assay providing a quantitative measure of the castration-resistant phenotype would help to guide clinical decisions related to the timing or type of treatments given to patients that are candidates for hormone therapies, thereby addressing a critical unmet need in the clinical management of advanced prostate cancer. Second, the detection assay can be applied to patients with progressive castration-resistant prostate cancer, and helps to identify a subset of patients that may continue to respond to second-line hormone therapies. Third, such AR variants and associated functions may not be targeted by any of the existing regimens or second-line hormone therapies currently under development (e.g., MDV3100 and abiraterone). As such, they are predicted to survive hormone ablation and propagate during the course of clinical progression. Novel therapies targeting these variants will have the potential to eliminate this subpopulation of pre-existing androgen-independent cells when applied in hormone-naive patients, and may be used to treat patients relapsed from first- and second-line hormone therapies. These future efforts focusing on both detection of modification of AR variants will go hand-in-hand, as development of new therapies would be greatly facilitated by the detection method that allows objective patient recruitment and real-time monitoring of clinical responses.

In summary, recent findings have shed light on an alternative mechanism explaining why prostate cancer patients fail hormone therapies. For the first time, the decoding and characterization of novel AR variants made it possible to detect and manipulate prostate cancer cells with constitutively active AR signaling under complete hormone ablation. Future studies will address the relative importance and clinical relevance of ligand-dependent versus ligand-independent routes toward hormone therapy failure and focus on the development of methods and approaches to detect and modify the ligand-independent AR-signaling pathway.

Financial & competing interests disclosure

Jun Luo was supported by the David H Koch Foundation for his research on advanced prostate cancer. Jun Luo is also a Phyllis and Brian L Harvey Scholar, awarded by the Patrick C Walsh Prostate Cancer Research Fund. The author has no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.

No writing assistance was utilized in the production of this manuscript.