Targeting androgen receptor signaling to enhance cancer immunotherapy.

Preclinical studies across multiple tumor types demonstrate that AR suppression reverses T cell exhaustion, enhances immune activation, and sensitizes tumors to ICIs. Emerging early-phase clinical trials suggest that these effects may extend beyond prostate cancer, providing a rationale for combining hormone modulation with immunotherapy in other malignancies. Optimizing the timing and duration of AR suppression will be critical to fully harness its immunologic and therapeutic potential. While most studies on the immunosuppressive effects of ADT have focused on PCa, growing evidence suggests that AR signaling shapes antitumor immunity across many tumor types. Indeed, AR activity inversely correlates with immune cell infiltration and immunotherapy response across multiple cancer lineages [ 97 ], suggesting that cancer cell extrinsic AR signaling within the tumor immune microenvironment may be broadly relevant beyond conventional hormone-responsive tumors. In a carcinogen-induced model of urothelial carcinoma, one study found that male mice developed more aggressive tumors than females and linked this finding directly to AR signaling [ 7 ]. The study leveraged the Four Core Genotype (FCG) model to separate the effects of sex chromosomes from hormones, demonstrating that this disparity was driven by androgens rather than chromosomal sex. The difference was absent in mice lacking T cells, indicating a T cell–dependent mechanism. Single-cell analyses revealed that male tumors were enriched for Tcf7 + progenitor-exhausted CD8 + T cells sustained by AR-mediated Tcf7 activation, whereas female CD8 + T cells without AR-driven Tcf7 activation more readily differentiated into functional effectors. Disrupting AR signaling through castration or CD8 + T cell–specific AR deletion reduced exhaustion, enhanced tumor control, and improved ICI responsiveness [ 7 ]. In agreement, another preclinical bladder cancer study found that the combination of AR suppression with either bacillus Calmette-Guerin (BCG) or anti-PD-1 treatment was more effective than either treatment alone in the MBT-2 syngeneic mouse model of bladder cancer [ 98 ]. Additional preclinical and translational studies support a mechanism by which AR signaling suppresses CD8 + T cell function across multiple tumor types. In syngeneic mouse models of melanoma and colon cancer, AR-driven epigenetic and transcriptional programs reduced both effector function and stem-like capacity of tumor-infiltrating CD8 + T cells, resulting in larger tumors in males than females in hosts with functional T cells [ 8 ]. In agreement with this mechanism, the effect was absent in immunocompromised mice, where tumor growth was similar between sexes. Castration alleviated T cell suppression and combining AR blockade with ICI synergistically restricted tumor growth [ 8 ]. In mouse models of melanoma, colon cancer, lung cancer, breast cancer, and prostate cancer, AR signaling inhibited NF-κB activity in T cells, reducing the transcription of pro-inflammatory cytokines such as TNF and IL-2, which in turn impaired T cell proliferation and effector differentiation, ultimately contributing to greater tumor burden in male mice [ 9 ]. AR suppression restored NF-κB signaling, reversed these transcriptional defects, and improved ICI efficacy. Extending these observations to humans, CD8 + T cells with elevated AR levels exhibited reduced expression of cytotoxic genes (e.g., GZMB, IFNG) in mCRPC tumor biopsies obtained from patients during ICI treatment, recapitulating the AR-mediated transcriptional repression and functional impairment in preclinical models. Using CRPC and sarcoma models, AR inhibition restored IFNγ production, improved CD8 + T cell function, and sensitized tumors to ICI [ 5 ]. Collectively, these findings demonstrate that suppressing AR signaling can prevent or reverse T cell exhaustion and enhance ICI efficacy across multiple cancer types. Recent studies are beginning to translate preclinical insights on hormonal manipulation and immune activation into prospective trials combining ADT with ICIs in tumor types beyond PCa ( Figure 3B ). In a phase I study, ADT (triptorelin) combined with nivolumab was tested in 14 men with melanoma who had progressed after prior ICI therapy (anti-PD-1 or anti-PD-L1 ± anti-CTLA4), achieving disease control in 50% of patients by iRECIST[ 99 ]. Building on the broader concept that sex hormone suppression may restore or enhance responsiveness to immune checkpoint blockade, a phase 1b/2 trial evaluated hormone suppression (leuprolide) combined with an ICI and aromatase inhibition (exemestane) in 14 patients with premenopausal ER + /HER2 − metastatic breast cancer that had progressed after one or two lines of hormonal therapy (most commonly tamoxifen), achieving its primary endpoint of progression-free survival. Similar to prostate cancer, hormone receptor positive breast cancer is generally regarded as immunologically ‘cold’. However, tumor RNA profiling suggested the combination of hormone suppression and immune checkpoint blockade increased immune cell infiltration and enhanced ICI responsiveness [ 100 ]. The objective response rate (ORR) of 35.7% was markedly higher than historical controls, such as KEYNOTE-028 where ORR was reported at 12% [ 101 ]. Leuprolide suppresses both androgen and estrogen production in women, and this study should be interpreted cautiously given the single-arm design and limited sample size. However, these findings are under prospective evaluation in NCT06225284 , a larger Phase II trial testing neoadjuvant chemotherapy ± ADT ± pembrolizumab in premenopausal women with early triple negative breast cancer. Together with preclinical evidence that hormonal manipulation can augment immune responses, these results provide a compelling rationale for advancing ADT-ICI combinations into randomized trials in cancers not traditionally considered hormone-sensitive. While clinical trials in this space are limited, NCT06512207 is enrolling men aged 60 and older with advanced non-small cell lung cancer to test the combination of ADT and pembrolizumab, NCT05327647 is testing the combination of the anti-androgen bicalutamide in combination with BCG immunotherapy in patients with bladder cancer, and NCT03942653 is testing ADT + pembrolizumab for advanced AR positive salivary gland carcinoma. While murine studies reveal parallels with human biology and early human trials combining hormone manipulations with immune checkpoint inhibitors show promise, only large randomized clinical trials can capture the complexity of human immune responses and phenotypic diversity needed for broader therapeutic success. In addition to sex, human trials combining hormone manipulations with immunotherapy must also consider age as a biological variable impacting both AR signaling and immunosenescence [ 49 , 102 ]. Moreover, ADT reduces lean body mass and promotes insulin resistance, hepatic steatosis, and chronic inflammation, all of which can negatively impact quality of life and alter pharmacologic and immune responses. Given uncertainty regarding the need for long-term ADT in cancers that are not thought to be intrinsically driven by AR signaling, the side effects of long term ADT should be carefully balanced against the expected immunomodulatory ICI priming benefits. Nonetheless, ADT mitigation strategies (e.g., cardiometabolic optimization, resistance training, diabetes and osteoporosis management) have enabled the safe deployment of AR-directed therapies in prostate cancer for decades, and early trials combining ADT with ICIs in other tumors have not reported new or concerning safety signals [ 99 , 100 ]. Future trial designs should optimize the timing of hormone manipulations to prime the tumor immune microenvironment for maximal anti-tumor response ( Figure 3C ). Human studies in prostate cancer assessing the impact of neoadjuvant ADT prior to prostatectomy suggest a robust tumor immune infiltrate is induced by about 14 days based on transcriptional profiling and IHC assessment [ 74 , 82 ]. Other trials in mCSPC have applied ADT for either 4 or 10 weeks prior to ICI treatment and likewise reported a robust tumor immune infiltrate based on single cell RNA sequencing [ 83 ]. It remains uncertain which timepoint corresponds with maximal anti-tumor immune activity and it will likely vary based on tumor origin, location, and other host factors. In any case, if immunotherapy is administered after the ‘warming’ effects of ADT have waned, its efficacy may be markedly reduced. Currently there are no noninvasive biomarkers to assess the impact of AR suppression on the immune composition of the TME. Therefore, while basic measures such as serum testosterone or cytokine levels provide a starting point, achieving real-time non-invasive assessment of the tumor immune microenvironment will require more sophisticated tools such as peripheral immune profiling using flow cytometry and Immuno-PETCT imaging [ 103 ] to measure CD8 + T cell tumor infiltration in response to hormone manipulations. To date, no human studies have specifically assessed the optimal timing or sequencing of ADT and ICI. Most preclinical studies have applied ADT concurrently with ICI [ 5 , 7 – 9 ] and no human studies have assessed how ADT impacts the tumor immune infiltrate in malignancies beyond prostate cancer. Three Phase III trials testing ICI after prolonged ADT in prostate cancer failed to show benefit[ 87 – 89 ], suggesting that checkpoint blockade may have been applied during a suboptimal immunologic window ( Figure 3C ). Based on the immune kinetics observed in prostate cancer, 2–4 weeks of ADT may represent a reasonable starting point that will likely prime the tumor immune microenvironment for a more robust response to ICI therapy in other cancers. Conversely, studies of therapeutic prostate cancer vaccines such as sipuleucel-T demonstrate that ADT can enhance vaccine-induced immunity when administered after (rather than before) immunization, highlighting that the optimal sequence of hormone manipulation is likely context-dependent [ 104 , 105 ]. Finally, the optimal duration of ADT when used as an immunologic primer remains uncertain. In clinical practice, ADT is often maintained indefinitely for men with metastatic prostate cancer whereas preclinical studies typically continue hormone suppression throughout the experimental period. Human data defining the duration required to sustain immune activation are not yet available.

Sex-based disparities in cancer incidence and outcomes reflect fundamental biological differences shaped in part by AR signaling. Collective evidence from recent studies suggests that while AR inhibition initially promotes antitumor immunity, sustained suppression leads to immunosuppressive remodeling of the tumor microenvironment. In retrospect, the disappointing lack of efficacy observed in recent Phase III trials combining AR inhibition with ICIs in prostate cancer may reflect these underlying immune dynamics. Deciphering how AR signaling dynamically regulates the tumor immune microenvironment will be essential for translating AR – immunotherapy combinations into effective clinical strategies in prostate cancer and beyond. Men consistently have higher cancer incidence (~1.5-fold) and mortality (~1.8-fold) than women across nearly all shared anatomic sites [ 10 ]. The most pronounced mortality differences are seen in cancers of the bladder, esophagus, head and neck, liver, kidney, and cutaneous melanoma [ 10 , 64 ] ( Figure 2C ). While often attributed to behavioral and environmental exposures [ 65 ], these sex-based differences persist after controlling for known risk factors and are frequently observed in genetically engineered animal models of cancer[ 66 – 68 ], implicating biological differences. One noteworthy biological difference between sexes is tumor mutation burden (TMB). Male tumors generally exhibit higher TMB than female tumors, particularly in male-predominant cancers such as bladder, kidney, liver, and melanoma[ 69 ]. However, TMB differences alone do not consistently align with clinical outcomes. For example, glioblastoma is more common and lethal in men despite a significantly higher TMB in women [ 69 , 70 ]. These findings are consistent with the premise that sex disparities in cancer arise from a complex interplay between environmental exposures, sex chromosomes (which encode genes that influence metabolism and immunity), DNA repair capacity, and hormonal signaling. Among these, the immunomodulatory effects of AR signaling are emerging as a compelling and therapeutically relevant factor capable of shaping adaptive anticancer immunity. A prevailing hypothesis is that AR-mediated immunomodulation contributes to sex-specific differences in cancer incidence and responses to immunotherapy. With respect to toxicity, a meta-analysis including 23,296 patients found that women treated with ICIs experienced a 49% greater risk of severe adverse events than men [ 71 ], suggesting stronger baseline immunity associated with reduced AR signaling may predispose women to immune-mediated toxicity. However, evidence for sex-specific differences in ICI efficacy is more nuanced. One meta-analysis encompassing 20 trials across eight cancer types and 11,351 patients found that men exhibited a lower hazard ratio for death after ICI treatment compared with women in 19 of 20 trials [ 72 ]. This finding supports the idea that higher AR signaling results in lower baseline anticancer immunity in men, increasing their responsiveness to immune-activating therapies. However, a subsequent meta-analysis including 23,760 patients across 37 trials and 12 cancer types found no overall difference in ICI benefit between sexes [ 73 ]. This apparent inconsistency between studies likely stems from variations in trial design and differences in the statistical methods employed. On balance, these findings suggest that differences in ICI benefit between sexes are context-specific rather than universal, reinforcing the need for studies to identify tumors with androgen-responsive microenvironments that may be therapeutically targeted to enhance ICI efficacy. The widespread use of androgen suppression in PCa made it the natural starting point for early investigations into how androgen manipulation shapes anticancer immune responses. In one study, Androgen deprivation therapy (ADT) using a GnRH antagonist before prostatectomy in men with PCa triggered a robust oligoclonal T cell infiltrate that appeared to target prostate tumor cells [ 74 ]. Similarly, castration in mice enabled prostate-specific T cells to break through immune tolerance, expand, and develop effector function in response to stimulation with an engineered PCa-restricted antigen [ 75 ]. Moreover, AR inhibition enhanced TH1 differentiation [ 58 ], sensitized PCa cells to T cell killing [ 76 ], and improved survival in a spontaneous mouse model of PCa when used in combination with a metastasis vaccine [ 77 ]. On the other hand, ADT promoted prostate tumor progression by recruiting M2-like tumor-associated macrophages (TAMs) with reduced anticancer cytotoxicity into the tumor microenvironment (TME) [ 78 , 79 ]. Initially, castration promoted a TH1-polarized infiltrate supportive of anticancer immunity, but this response gradually shifted toward an immunosuppressive TH17 -driven TME with associated Treg expansion [ 80 , 81 ]. Thus, while ADT can trigger robust antitumor T cell responses and enhance immunotherapy efficacy in preclinical models, these benefits may be blunted by the recruitment of TAMs and other regulatory immune populations and likely diminish over time as the TME evolves. A series of recent preclinical and translational studies leveraged genomic and single-cell technologies to map how ADT remodels the prostate TME, revealing both its capacity to trigger rapid immune activation and its tendency to transition toward immunosuppression over time. For example, ADT using degarelix in men with localized PCa prior to prostatectomy resulted in a rapid increase in activated CD8 + T cells, suppressive Tregs, and pro-inflammatory M1-like tumor-associated macrophages in the TME [ 82 ]. ADT also altered tumor cells, which subsequently upregulated class I and II major histocompatibility complex (MHC) expression and downregulated the anti-phagocytic cell surface protein CD47. These combined effects may facilitate phagocytosis of tumor cells by nearby macrophages and facilitate T cell priming [ 82 ]. Similarly, AR inhibition increases MHC I expression in PCa cell lines and murine models, though this increase was transient, fading over time with the emergence of resistance to AR inhibition [ 6 ]. In men with metastatic hormone-sensitive prostate cancer (mHSPC), single-cell RNA sequencing demonstrated that ICI combined with ADT (following 4 weeks of ADT pretreatment) produced a more robust immune infiltrate including CD8 + and CD4 + T cells , as well as Tregs within metastatic sites compared to ADT alone [ 83 ]. Collectively, these studies demonstrate that ADT can initiate a short-lived immunogenic phase marked by enhanced antigen presentation, influx of activated T cells, and pro-inflammatory macrophage polarization, potentially amplifying the effects of immune checkpoint inhibition. However, this window of opportunity appears to close with sustained AR suppression, and the TME shifts toward immunosuppression with increased regulatory cell populations, reduced MHC expression, and impaired T cell priming ( Figure 2C ). Without strategies to optimize treatment timing or sustain immune activation, this ‘cooling-off’ phase may limit the clinical efficacy of combining AR suppression with ICIs in prostate cancer. Compared with other cancer types in which ICIs have proven effective, untreated PCa and mCRPC exhibit a sparse immune infiltrate [ 82 , 84 ], and ICIs have shown benefit only in prostate tumors with microsatellite instability-high (MSI-H) or mismatch repair deficient (dMMR) disease. Nonetheless, preclinical evidence and early-phase clinical data [ 85 , 86 ] provided a strong rationale for combining ADT with ICIs to enhance antitumor responses in patients with biomarker-unselected PCa. KEYNOTE-991 was a randomized, double-blind phase III trial that tested the addition of pembrolizumab to ADT and the anti-androgen enzalutamide in 626 men with mHSPC who had been treated with ≤ three months of ADT or six or fewer cycles of docetaxel with ADT prior to enrollment. This trial found no difference in radiographic progression-free survival nor median overall survival at the first interim analysis and so was stopped for futility [ 87 ]. In addition, two recent phase III studies (IMbassador250 and KEYNOTE-641) tested the addition of ICI to enzalutamide in men with mCRPC who had progressed on ADT, and in some cases abiraterone or docetaxel prior to enrollment. Both trials found no improvement in overall survival compared with enzalutamide alone [ 88 , 89 ]. Preplanned subgroup analysis in IMbassador250 identified longer progression free survival with atezolizumab and enzalutamide in patients with higher PD-L1 immune cell expression, higher CD8 + T cell density, and increased T-effector signatures, indicating that the immune phenotype predictive of ICI benefit occurs in prostate cancer, but in relatively few patients [ 88 ]. Prespecified subgroup analyses in KEYNOTE-641 and KEYNOTE-991 did not identify any molecularly defined subgroups with clear benefit. These trials did not mandate on-treatment biopsies and so molecular evaluation of the impact of ADT in combination with ICI on the tumor immune microenvironment in these trials is limited. Although the molecular basis for the general lack of efficacy remains uncertain, several factors may plausibly account for these results. First, prolonged exposure to continuous ADT, often for months or years before trial entry, can generate an immunosuppressive TME, and there is no evidence that adding enzalutamide after this degree of suppression restores immune function or primes the TME for ICI responsiveness. Second, AR antagonists such as enzalutamide can directly suppress T cell activation via off-target activation of the GABA-A receptor [ 90 ] and, as demonstrated in preclinical models of prostate and colon cancer [ 91 , 92 ], can paradoxically promote immune-mediated tumor progression by enhancing myeloid-derived suppressor cell (MDSC)- mediated suppression of T cells. Third, primary resistance to enzalutamide, often mediated by persistent AR signaling in tumor parenchymal cells, occurs in ~20% of patients with mCRPC [ 93 ], potentially overwhelming any immune-mediated benefit of ICI. Together, prolonged ADT-induced immunosuppression, direct and indirect immunosuppressive effects of enzalutamide, and the prevalence of primary resistance could have limited the efficacy of ICI combinations observed in these trials. Future studies seeking to combine ADT with ICI therapy in PCa may benefit from selecting alternative AR-targeting agents, intermittently restoring AR signaling, and timing ICI delivery to coincide with maximal ADT-induced immune activation ( Figure 3A ). For example, instead of enzalutamide, future studies should select AR inhibitors that are not known to impair T cell activation, such as abiraterone acetate or darolutamide. While combined AR blockade (ADT plus a second agent) is probably necessary to suppress tumor-intrinsic AR signaling in mCRPC, early-phase data from the phase 1b/2 KEYNOTE-365 Cohort D suggest that abiraterone plus ICI is both feasible and potentially active [ 94 ]. Next, bipolar androgen therapy (BAT) , the intermittent administration of supraphysiologic testosterone, can reverse ADT-induced immunosuppression and re-establish pro-inflammatory changes within the TME, potentially sensitizing tumors to ICI [ 95 ]. Similarly, another approach is to initiate ICIs during the transient immunogenic “warming” of the TME that follows ADT. These latter strategies, aimed at immunologically priming the TME through androgen modulation, may be particularly important in PCa given its generally immunologically “cold” phenotype characterized by low tumor mutational burden and limited neoantigen load [ 84 , 96 ]. Current clinical trials testing combination of AR suppression with ICIs in prostate cancer include NCT07027124 , which will test neoadjuvant ADT, darolutamide, and pembrolizumab prior to radical prostatectomy in men with high risk nonmetastatic prostate cancer, and NCT04946370 will test the combination of the prostate specific membrane antigen (PSMA) conjugated alpha emitter 225Ac-J591 with an antiandrogen (enzalutamide, apalutamide, or darolutamide) and pembrolizumab. Another trial NCT04221542 will combine abiraterone or enzalutamide with the bispecific T cell engager xaluritamig (anti-CD3/anti-STEAP1) in men with metastatic prostate cancer.

Accumulating evidence defines the critical role of androgens in shaping anticancer immunity, and harnessing this biology is expected to enable more effective, sex-tailored cancer treatments. Preclinical studies and early clinical trials demonstrate that suppressing AR signaling enhances anticancer immune responses across multiple tumor types, offering a potential avenue to overcome immune resistance. Translating these findings to clinical practice will require clinical trials that integrate key biological variables including sex, age, and immune state, while carefully balancing the immunologic benefits of ADT against long-term metabolic and quality-of-life effects. Future trial designs should time AR suppression to coincide with peak immune activation, define the optimal duration of ADT, and explore ADT-ICI combinations in tumors not traditionally considered to be hormone responsive (see Outstanding Questions ).