Prostate cancer is the fifth leading cause of cancer death worldwide. At the time of diagnosis, approximately 80 to 90 percent of cases depend on male hormones called androgens to grow, making androgen blockade the cornerstone of treatment.
The primary androgens are testosterone and its more potent form, 5-alpha-dihydrotestosterone (DHT). These hormones work by binding to the androgen receptor (AR), a protein inside prostate cells. Once activated, the AR moves into the cell nucleus and switches on genes that promote cell survival and proliferation.
Beyond its nuclear role, the AR can also act rapidly outside the nucleus by interacting with signaling molecules that control the cell cycle, cell movement, and other growth-related processes. This dual activity makes the AR a complex and multi-layered target for therapy.
Despite initial success with androgen-blocking therapy, 20 to 30 percent of patients progress to castration-resistant prostate cancer (CRPCa) -- a state where the tumor continues to grow even when testosterone is reduced to very low levels. Understanding why this happens is central to improving long-term outcomes.
The androgen receptor gene encodes a protein with three key functional regions: the N-terminal transactivation domain which activates gene transcription, the DNA-binding domain (DBD) which attaches to specific DNA sequences, and the androgen-binding domain (ABD) where testosterone and DHT attach. Each domain is encoded by specific regions of the gene.
AR is found in nearly all primary and metastatic prostate cancers. Even in tumors that no longer respond to hormone therapy, AR expression is usually maintained, suggesting the cancer has found alternative ways to keep the receptor active rather than abandoning it altogether.
AR is also expressed in the stromal cells -- the supportive non-cancerous tissue surrounding the tumor. Interestingly, low stromal AR is linked to worse outcomes, while high AR in the cancer cells themselves shows inconsistent prognostic value. As the cancer becomes more aggressive and dedifferentiated, AR expression tends to drop in both compartments.
During cancer progression, tumor cells shift from relying on signals from surrounding stromal tissue to producing their own growth signals internally. This shift from paracrine to autocrine signaling represents a key step toward independence from normal tissue controls.
AR mutations are rare in untreated prostate cancer (about 1%) but become common in metastatic disease (up to 60%). Many mutations occur in the androgen-binding domain and cause a critical problem: they allow drugs designed to block the receptor to instead activate it, turning an intended inhibitor into a fuel source for the tumor.
One of the most studied mutations, T877A, is found in about 30% of metastatic castration-resistant prostate cancer cases. It broadens the range of hormones that can activate the receptor, including weak adrenal androgens that are not suppressed by standard castration therapy. Other mutations confer resistance to specific modern drugs like enzalutamide and apalutamide.
AR gene amplification -- having extra copies of the gene -- occurs in 20 to 30% of treated and castration-resistant cases. Extra gene copies increase the receptor's sensitivity, allowing the cancer to grow at androgen levels too low to drive normal cells. This is another mechanism by which prostate cancer escapes hormone therapy.
AR splice variants are altered forms of the AR protein produced when parts of the gene's instructions are skipped during processing. The most clinically important is AR-V7, which lacks the androgen-binding domain entirely. This means AR-V7 is permanently switched on and cannot be blocked by drugs that target the binding domain. Patients whose tumors express AR-V7 tend to respond poorly to standard hormone therapies and have shorter survival.
AR gene amplifications can be detected in tumor tissue or in circulating tumor cells (CTCs) using a technique called fluorescence in situ hybridization (FISH). Detection in blood-based circulating DNA has proven strongly prognostic across multiple clinical studies, making it a promising non-invasive biomarker.
AR-V7 detection is achievable from multiple sample types including biopsies, CTCs, plasma, and even whole blood. A key study using a highly specific RNA-based detection method showed that higher AR-V7 expression in tumor biopsies correlated with poorer response to the drugs abiraterone and enzalutamide in metastatic castration-resistant disease.
An immunohistochemistry-based CTC test -- the Oncotype Dx AR-V7 Nucleus Detect Test -- can identify which patients are more likely to benefit from chemotherapy over hormone-targeted drugs. Patients with AR-V7 positive circulating tumor cells responded better to taxane chemotherapy than to further androgen receptor inhibition. This test has received coverage authorization in the United States.
Despite these advances, the overall evidence is still not strong enough for routine systematic testing according to major pathology organizations. The prevalence of AR-V7 in castration-resistant populations varies widely across studies (from 19% to 70%), partly due to differences in detection methods, which complicates standardization and clinical adoption.
Beyond AR mutations and splice variants, tumors can escape treatment through other AR-dependent and AR-independent mechanisms. One involves intratumoral androgen synthesis -- the cancer making its own supply of androgen from precursor molecules, bypassing the external supply that standard castration removes.
An AR-independent mechanism called lineage switching occurs when prostate cancer cells, under sustained pressure from AR-targeted drugs, transform into a completely different cell type. Approximately 25% of metastatic prostate cancer patients develop neuroendocrine differentiation, in which tumor cells acquire stem-like properties and low or absent AR expression. This phenotype is associated with very poor outcomes.
Bypass signaling is another AR-independent escape route. In this scenario, a signaling molecule outside the AR pathway activates the same downstream genes that AR normally controls. One driver of this is believed to be enhanced activity of the glucocorticoid receptor, which can substitute for AR in driving gene expression programs that support cancer survival.
These diverse resistance mechanisms highlight why no single therapeutic strategy can cure advanced prostate cancer. The tumor's genetic flexibility allows it to adapt around any single point of attack, motivating ongoing research into combination treatments and sequential therapy strategies.
Three new androgen receptor signaling inhibitors (ARSi) -- enzalutamide, apalutamide, and darolutamide -- have each demonstrated significant benefits in patients with non-metastatic castration-resistant prostate cancer (men whose PSA is rising rapidly but whose cancer has not yet spread to other organs on imaging).
In the phase 3 PROSPER trial, enzalutamide extended the median time before metastasis or death from 14 months with placebo to 36 months, a 71% reduction in risk. It also improved overall survival by over 10 months compared to placebo and delayed the need for additional cancer treatment by about two years.
The SPARTAN trial showed that apalutamide extended metastasis-free survival from 16 to 40 months, reducing the risk of metastasis or death by 72%. The ARAMIS trial demonstrated that darolutamide reduced the risk of death by 31% compared to placebo, with benefits seen across multiple secondary endpoints including time to pain progression and skeletal events.
Notably, darolutamide has a distinct chemical structure that does not cross the blood-brain barrier as readily as enzalutamide, potentially explaining its lower rates of central nervous system side effects such as fatigue and cognitive impairment. Updated data from all three trials presented in 2020 confirmed overall survival benefits for all three agents.
The success of these agents in castration-resistant disease prompted clinical trials in castration-sensitive prostate cancer -- patients who still respond to standard androgen deprivation therapy but have high-risk features. Adding abiraterone to standard therapy in the LATITUDE trial extended median overall survival from 36 to 53 months in high-risk metastatic patients.
The ARCHES and ENZAMET trials showed that adding enzalutamide to standard androgen deprivation therapy significantly reduced the risk of radiographic progression or death compared to hormone therapy alone, establishing enzalutamide as a new option in this setting. Benefits were seen even in patients with low-volume disease and those who had already received docetaxel chemotherapy.
The TITAN trial demonstrated that apalutamide plus standard therapy improved both overall survival and radiographic progression-free survival in metastatic castration-sensitive patients compared to hormone therapy alone, with similar rates of serious adverse events in both treatment arms.
An important emerging issue is that as more patients receive potent AR-targeted therapies early in their disease course, a larger proportion may subsequently develop AR-independent tumors with neuroendocrine features. Managing this future population of patients with treatment-emergent neuroendocrine prostate cancer represents a major clinical challenge ahead.
Researchers are testing combinations of AR inhibitors with drugs from other classes to overcome resistance. PARP inhibitors (PARPi) such as niraparib, which block DNA repair, are being combined with ARSi agents in clinical trials for patients with or without DNA damage repair gene mutations. Early data suggest benefit particularly in patients whose tumors harbor homologous repair defects.
Immunotherapy combinations are also under investigation. However, the IMbassador250 trial showed no survival benefit from adding the checkpoint inhibitor atezolizumab to enzalutamide after prior abiraterone failure, suggesting that prostate cancer remains a difficult target for standard immune checkpoint blockade without additional biomarker selection.
A practical concern raised by combination studies is drug-drug interactions. A phase 1 trial found that enzalutamide dramatically reduces blood levels of the c-MET inhibitor crizotinib by inducing a liver enzyme that breaks down the drug faster. This highlights the need for pharmacokinetic studies before combining AR agents with other targeted therapies.
The field is moving toward biomarker-driven treatment selection. Established examples include olaparib for patients with homologous repair gene mutations and pembrolizumab for tumors with microsatellite instability. For the majority of patients, better predictive biomarkers that identify who will respond to which AR-targeted therapy are urgently needed.