Clinical impact of real‐time androgen receptor alteration monitoring on metastatic castration‐resistant prostate cancer treatment in real‐world settings

Abstract

Androgen receptor (AR) alterations contribute to resistance against androgen receptor signaling inhibitors (ARSi) in metastatic castration‐resistant prostate cancer (mCRPC). This study evaluated AR alteration monitoring via liquid biopsy in routine clinical practice. To this end, we enrolled 39 mCRPC patients in a real‐world clinical setting and monitored disease progression with progression‐free survival (PFS) and overall survival (OS) analyzed in relation to AR status. AR alterations were detected in 8 of 39 patients (20.5%) at baseline, with five additional cases emerging during progression (total 33.3%). AR‐V7 was identified in 12.8%, and AR amplification and/or hotspot mutations in 20.5%. Patients with AR alterations had significantly lower PSA response rates to ARSi (37.5% vs. 80.7%; p = 0.0276). All AR alteration‐positive patients experienced disease progression, compared to 34.6% of AR‐negative cases. PFS was significantly shorter in AR alteration‐positive patients (11 vs. 52 months; p = 0.001), while OS showed a non‐significant trend toward shorter survival (41 vs. 74 months; p = 0.0619). Univariate analysis confirmed AR alterations as an independent predictor of PFS (p = 0.0035). This real‐world study demonstrates that AR monitoring via liquid biopsy predicts treatment response and progression in mCRPC. Continuous monitoring is essential, as AR alterations emerge over time. Patients with AR alterations have poorer ARSi responses and shorter PFS, emphasizing the need for adaptive treatment strategies in real‐world clinical practice.

What's New?

Androgen receptor (AR) alterations contribute to treatment resistance against androgen receptor signaling inhibitors in metastatic castration‐resistant prostate cancer. This real‐world study demonstrates that AR monitoring via liquid biopsy predicts treatment response and progression in metastatic castration‐resistant prostate cancer. It highlights the dynamic nature of AR alterations, their role in resistance to AR inhibitors, and their association with poor outcomes. Patients with AR aberrations may benefit more from chemotherapy, emphasizing the need for predictive biomarkers. Liquid biopsy may serve as a non‐invasive, cost‐effective tool for routine real‐time AR monitoring, enabling adaptive cancer therapy in metastatic castration‐resistant prostate cancer in clinical practice.

Abbreviations

AA

Abiraterone acetate

ADT

Androgen deprivation therapy

APA

Apalutamide

AR

Androgen receptor

ARSi

Androgen receptor signaling inhibitors

AR‐FL

Androgen receptor full‐length transcript

AR‐V7

Androgen receptor splice variant 7

CAB

Cabazitaxel

cfDNA

Cell‐free DNA

CNA

Copy number alteration

CRPC

Castration‐resistant prostate cancer

CT

Computed tomography

DARO

Darolutamide

DNA

Deoxyribonucleic acid

DOC

Docetaxel

ENZ

Enzalutamide

HSPC

Hormone‐sensitive prostate cancer

IV

Intravenous

KLK3

Kallikrein‐related peptidase 3 (PSA gene)

LDH

Lactate dehydrogenase

MA

Massachusetts

mCRPC

Metastatic castration‐resistant prostate cancer

NE

Nebraska

OS

Overall survival

PCR

Polymerase chain reaction

PFS

Progression‐free survival

PSA

Prostate‐specific antigen

PSA50

≥50% reduction in PSA levels

RECIST

Response evaluation criteria in solid tumors

RLT

Radioligand therapy

RNA

Ribonucleic acid

RPE

Radical prostatectomy

SNV

Single nucleotide variant

1. INTRODUCTION

Prostate cancer is the most commonly diagnosed cancer in men and the second leading cause of cancer‐related deaths in the Western population. ¹ , ² While patients diagnosed with localized prostate cancer have excellent 5‐year survival rates, the median overall survival (OS) of metastatic prostate cancer is only 3–5 years. ³ As the androgen receptor (AR) plays an important role in the progression of prostate cancer, mediating the effects of androgens on cellular proliferation and survival, ⁴ a systemic lowering of androgen levels through androgen deprivation therapy (ADT) is the first‐line treatment. While ADT is initially effective in controlling hormone‐sensitive prostate cancer (HSPC), most patients eventually develop resistance, leading to castration‐resistant prostate cancer (mCRPC), which further declines OS to 1–3 years, depending on prior treatments and biomarker status. For patients with mCRPC, the next therapeutic step after ADT failure is the use of androgen receptor signaling inhibitors (ARSi), which more effectively block AR activity. ⁵ These therapies have significantly improved survival in mCRPC, ⁶ , ⁷ yet resistance to ARSi remains a major challenge, limiting long‐term effectiveness.

Recent studies have demonstrated that alterations in the AR—such as amplifications, splice variants (e.g., AR‐V7), and point mutations—are strongly associated with resistance to ARSi. ⁸ , ⁹ , ¹⁰ , ¹¹ , ¹² Among these, AR‐V7 has emerged as a particularly strong predictive biomarker for ARSi resistance, while AR amplifications and point mutations‐especially in the ligand‐binding domain—may contribute to variable levels of resistance and are often associated with poorer clinical outcomes. Some point mutations within this domain are known to mediate resistance by converting antagonists like enzalutamide into partial agonists, while others broaden the receptor's ligand specificity, allowing activation by alternative steroids such as progesterone or by glucocorticoids like prednisone. ⁸ , ⁹ , ¹⁰ , ¹² , ¹³ , ¹⁴ , ¹⁵ , ¹⁶

Identifying these alterations may allow for better treatment selection and optimization of therapeutic sequences. Previously, we have developed and validated five multiplex droplet digital PCR (ddPCR) assays for detecting AR amplification, AR‐V7, KLK3 expression, and AR point mutations from circulating nucleic acids derived from a liquid biopsy. Unlike traditional tissue biopsies, liquid biopsies can provide a comprehensive snapshot of tumor heterogeneity and dynamics over the course of treatment. Several studies have demonstrated the feasibility of detecting AR alterations in liquid biopsy samples and have correlated these alterations with clinical outcomes in mCRPC patients. ¹³ , ¹⁷ , ¹⁸ , ¹⁹

Our assays demonstrated excellent analytical sensitivity and specificity, confirming their potential utility in clinical practice. Consistent with other studies, our findings indicated that AR alterations were significantly more prevalent in patients with CRPC compared to those in the hormone‐sensitive stage. ¹³ , ¹⁷ , ¹⁸ , ¹⁹ Building on our previous findings, this study evaluates the clinical utility of AR alteration monitoring in a real‐world cohort of mCRPC patients using liquid biopsy. Unlike controlled clinical trials, this observational study provides real‐world evidence from patients receiving standard‐of‐care treatments, reflecting actual clinical practice. By analyzing cfDNA and cfRNA from 39 patients, we assessed the prevalence and evolution of AR alterations and their impact on treatment response and survival outcomes. This research contributes to the growing body of evidence supporting the integration of liquid biopsy into routine patient management, enabling personalized and adaptive treatment strategies for advanced prostate cancer in real‐world settings.

2. MATERIALS AND METHODS

2.1. Patients and Samples

A total of 103 plasma samples were analyzed from 39 men diagnosed with mCRPC, who were at various stages of treatment at the Salzkammergutklinikum between July 2022 and June 2024. All patients had a histologically confirmed diagnosis of prostate adenocarcinoma and were classified as having CRPC according to the guidelines of the European Association of Urology. Blood samples were collected from patients before therapy initiation and/or during treatment as part of routine clinical care, with collection timing determined by the attending physician's discretion to gather additional information for treatment decisions. No specific study protocol or standardized guidelines dictated the timing of sample collection. Up to four follow‐up assessments were conducted to monitor AR alteration status throughout the disease course, with treatment strategies adjusted based on AR alteration findings.

2.2. cfDNA and cfRNA extraction from plasma and ddPCR analysis

A minimum of 20 mL peripheral blood was collected from each patient into three cfRNA blood collection tubes (Streck, Omaha, NE). Plasma was extracted within 72 h using a double spin protocol (1800×g for 15 min, 2800×g for 15 min) at the Department of Pathology at Salzkammergutklinikum as recommended by the manufacturer.

Total nucleic acids, including cfDNA and cfRNA, were isolated from 10 mL of plasma (2 × 5 mL) using the QIAamp Circulating Nucleic Acid Kit (QIAGEN), according to the manufacturer's instructions, and eluted in 60 μL of elution buffer. Plasma‐derived cfRNA was extracted without adding carrier RNA as previously described. ²⁰ The total nucleic acid eluates were purified further with the RNeasy MinElute Cleanup Kit (QIAGEN) and eluted in 14 μL of RNase‐free water. Of the 14 μL eluted cfRNA, 11 μL was reverse transcribed using a modified reverse transcription protocol with SuperScript IV reverse transcriptase (Thermo Fisher Scientific, Waltham, MA).

Five multiplex ddPCR assays were used for the detection of 13 AR targets including AR amplification [copy number alteration (CNA)], AR‐V7, KLK3, and a total of 10 AR hotspot mutations in codons 702, 742, 875, 877, and 878. The CNA assay was designed using primers and probes targeting the AR gene, along with three validated reference genes: ARHGEF9, NSUN3, and AGO1. For cfRNA gene expression targets—AR‐FL, AR‐V7, and KLK3—HPRT1 served as the endogenous control. Each reaction included no‐template controls and positive controls with known target profiles to ensure assay performance. Due to the design of the hotspot mutation assays, mutations at codons 877 and 878, as well as 702 and 742, cannot be distinguished individually. Consequently, results from these regions are reported as AR p.T878A/S‐F877L and AR p.L702H‐W742L/C, respectively. The details, development, and performance characteristics of all five assays have been thoroughly described previously. ²⁰

2.3. Statistical analysis

Kaplan–Meier method was used to evaluate time‐to‐event outcomes, including progression‐free survival (PFS) and overall survival (OS). PFS was defined as the interval from the initiation of the current ARSi therapy, or from the start of the most recent ARSi therapy if treatment had already concluded, until documented disease progression or death. OS was measured from the initiation of the first ARSi therapy to the time of death from any cause. Survival differences were compared by the log‐rank test.

Biochemical progression was defined as a ≥25% increase and an absolute rise of ≥2 ng/mL above the nadir, confirmed by a second consecutive measurement at least 3 weeks later. Clinical and/or radiographic progression was defined as either symptomatic progression (worsening symptoms related to the disease), radiologic progression (≥20% increase in the sum of diameters of soft‐tissue target lesions on CT scan, according to RECIST [Response Evaluation Criteria in Solid Tumors]), the appearance of two or more new bone lesions on bone scan, or death—whichever occurred first.

The PSA response rate was calculated according to the Prostate Cancer Clinical Trials Working Group 2 (PCWG2) criteria, which define a PSA response as a decline of ≥50% from baseline, confirmed by a second measurement at least 3 weeks later. In our study, baseline was defined as the initiation of first‐ or second‐line ARSi therapy.

The statistical significance of the difference in PSA response rates between AR‐alteration positive and negative groups was assessed using Fisher's exact test, with significance defined as a p‐value <0.05. Categorical variables, such as Gleason score at diagnosis, presence of bone, lymph node, and visceral metastases, and type of hormonal treatments, were described by absolute and relative frequencies.

All statistical analyses were performed using GraphPad Prism version 10.4.0 (GraphPad Software, Inc., La Jolla, CA).

3. RESULTS

3.1. Patient characteristics and timing of blood sampling

The primary objective of this study was to evaluate the clinical utility of AR alteration monitoring in routine oncology practice. Between July 2022 and June 2024, we enrolled 39 patients with CRPC with a median survival of 59 months and a median follow‐up time of 52 months (range, 4–100) and monitored their progress through September 2024. A subset of patients (n = 19) was previously described in Stitz et al. ²⁰ Patients' median age was 77 years (range, 60–87); Bone and/or lymph node metastases were observed in all patients (100%), while only one patient (2.6%) had additional visceral metastases. Patient characteristics are summarized in Table 1. Using real‐world clinical data, we assessed PFS and OS in relation to AR status. A secondary objective was to investigate the detectability and temporal dynamics of AR alterations in response to ARSi and chemotherapy. However, due to the limited number of CRPC patients within the cohort, AR testing prior to the initiation of the first ARSi treatment was only feasible in a subset of patients (n = 6). The majority of patients had already initiated ARSi therapy at the time of their first AR assessment, thereby influencing the ability to evaluate AR alterations in an ARSi treatment‐naïve context.

TABLE 1.

Study population characteristics presented as mean (range) or number (percentage).

	Total (n = 39)
Patients (CRPC)	39
Mean age (years)	77 (60–87)
Gleason score (%)
≤7	10 (25.6)
>7	26 (66.7)
N.A.	3 (7.7)
Metastatic status (%)
Bone metastases	32 (82.2)
Lymph metastases	23 (59.0)
Visceral metastases	1 (2.6)
Definitive therapy (%)
Radical Prostatectomy	10 (25.6)
Radiotherapy	28 (71.8)
Both	6 (15.4)
None	7 (17.9)
Systemic therapy (prior and/or current use)
ADT	38 (97.4)
CYP17A inhibitor (Orteronel)	1 (2.6)
Chemotherapy	14 (35.9)
None	0 (0.0)
Novel antiandrogens (prior or current use)
Abiraterone (AA) only	17 (43.6)
Enzalutamide (ENZ) only	8 (20.6)
Apalutamide (APA) only	3 (7.7)
Darolutamide (DARO) only	1 (2.6)
APA followed by ENZA and/or AA	5 (12.8)
ENZA followed by AA	3 (7.7)
AA followed by APA	1 (2.6)
DARO followed by APA and ENZA	1 (2.6)

Abbreviations: AA, abiraterone; ADT, androgen deprivation therapy; APA, apalutamide; DARO, darolutamide; ENZ, enzalutamide; N.A., not available.

At the time of the first blood draw, 10 patients had not yet started their first or second line ARSi therapy. A total of 23 patients were receiving ongoing ARSi treatment (abiraterone, n = 7; enzalutamide, n = 8; apalutamide, n = 7; darolutamide, n = 1), while six patients received either non‐ARSi therapies (e.g., chemotherapy, radioligand therapy) or no systemic treatment.

Within the entire cohort, 22 exhibited biochemical, clinical, and/or radiographic progression, while 17 patients did not experience disease progression during the study period. Of those who progressed, 15 patients underwent AR testing at or after their first progression event, while seven patients were tested prior to disease progression. The treatment regimen, progression status, PSA response (defined as a ≥50% reduction in PSA levels, PSA50), along with the timing and results of AR‐status assessment are presented in Figure 1.

FIGURE 1.

Swimmer plot by ARSi treatment, timing and results of AR status determination, progression and response. Each lane is color‐coded based on whether the patient received ARSi therapy (light blue with different patterns), chemotherapy (pink), radioligand therapy (olive) or no systemic therapy (gray). Orange and red vertical bars indicate the timing and result of the measurement (AR status negative vs. AR status positive). Diamonds indicate PSA response >50% (green), PSA progression (yellow) or death (black). ADT, Androgen Deprivation Therapy; CAB, cabazitaxel; DOC, docetaxel; RPE, radical prostatectomy; RLT, radioligand therapy.

3.2. Plasma AR status and PSA response rate

Considering only the first blood draw, AR alterations were detected in 8 of 39 patients (20.5%) (Figure 2B). During the observation period, five additional patients, who initially had AR wild‐type status, subsequently presented detectable AR alterations in a follow‐up sample (Figures 3A and S1A–D). Therefore, altogether AR alterations were identified in 13 of 39 patients (33.3%) (Figure 2C). Among these, the AR‐V7 splice variant was identified in five patients (12.8%), while AR amplification and/or AR hotspot mutations were observed in eight patients (20.5%). Furthermore, seven of those patients (54%) harbored more than one AR alteration, with five patients exhibiting a combination of AR‐V7, AR amplification, and/or hotspot mutations, and two patients displaying multiple AR hotspot mutations. In total, we identified 22 AR altering events across the 13 patients: five AR‐V7 results, seven AR gene amplifications findings, and 10 AR hotspot mutation detections (Figure 2C). Moreover, 12 patients (92.3%) also had detectable levels of KLK3, while all tested patients exhibited detectable full‐length AR (AR‐FL) transcript levels confirming the presence of amplifiable and intact cfRNA from prostate‐derived tissue in liquid biopsy samples (Figure 2C).

FIGURE 2.

Waterfall plot of best prostate‐specific antigen (PSA) responses by AR alteration status and heatmaps of detected AR alterations at baseline and at last follow‐up. (A) Waterfall plot showing PSA change (%) on the Y‐axis and individual patients on the X‐axis. Asterisks (*), triangles (∆), and circles (○) indicate the timepoint of AR testing: At baseline (n = 10), <7 months after ARSi initiation (n = 10), and >7 months after ARSi initiation (n = 19), respectively. Bar colors represent AR alteration status: Gray for AR alteration‐negative, dark orange for AR‐V7 positive alone or in combination with AR amplification (CNA) and/or AR hotspot mutations (SNV), and light orange for AR CNA and/or AR SNV positive. (B, C) Heatmaps display cfRNA markers including AR‐FL, KLK3, AR‐V7, as well as AR amplification and hotspot mutations at codons 702, 742, 875, 877, and 878. The orange to black gradient indicates expression or mutation intensity; gray indicates absence of the target. (B) AR alterations at the first blood draw, (C) AR alterations at the last follow‐up. AR, androgen receptor; cfRNA, cell‐free RNA; AR CNA, AR amplification; AR SNV, AR hotspot mutations; AR‐FL, androgen receptor full‐length; KLK3, kallikrein‐related peptidase 3; AR‐V7, androgen receptor splice variant 7.

FIGURE 3.

A representative panel of individual PSA progression in AR alteration‐positive and ‐negative patients, along with a comparison of PSA change rates between the two groups. Panels A–D display individual PSA trajectories on a log10 scale, spanning from the initiation of the first ARSi therapy to either the end of the study observation period or the patient's death. Each panel indicates the AR status and highlights key treatment interventions. Treatments are represented as ARSi therapy (light blue with various patterns), chemotherapy (pink), radioligand therapy (olive), or no systemic therapy (gray). PSA trajectories for all patients are shown in Figures S1, S2 and S3. Panels E and F provide a grouped comparison of PSA change rates (%) between 26 AR alteration‐negative and 13 AR alteration‐positive patients, with the PSA change rate represented on the y‐axis. ADT, androgen deprivation therapy; AR, androgen receptor; ARSi, androgen signaling inhibitor therapie; SNVs*, AR p.T878A/S‐F877L + AR p.L702H‐W742L/C.

To assess the predictive value of AR alteration screening, we calculated the PSA response rates of all 39 patients according to their AR alteration status (Figure 2A). Six patients were receiving either non‐ARSi therapy or no systemic treatment at the time of sample collection; however, all had previously received abiraterone, with three patients demonstrating no therapeutic response to abiraterone.

Among the 39 mCRPC patients analyzed, 28 (71.8%) exhibited a PSA decline of >50% following treatment. However, the PSA response rate to ARSi therapy was significantly lower in patients with AR alterations, with only 37.5% (3 of 8 patients) responding, compared to 80.7% (25 of 31 patients) in AR alteration‐negative cases (Fisher's exact test, p = 0.0276) (Figure 2A). When considering only AR‐V7, the difference in response to treatment is even more significant. None of the patients with AR‐V7 expression (0 of 4) showed a PSA response, whereas 80% of AR‐V7 negative patients (28 of 35) responded to therapy (Fisher's exact test, p = 0.004). In contrast, patients with AR amplifications or hotspot mutations demonstrated lower response rates compared to negative individuals, although these differences were not statistically significant. Response rates were 50% in patients with AR amplification versus 74.3% in those without (p = 0.5619), and 60% versus 73.5% in patients with and without AR hotspot mutations, respectively (p = 0.6088). These findings should be interpreted with caution, as not all patients were tested for AR alterations at baseline. Consequently, some patients classified as AR alteration‐positive may have acquired these alterations during therapy, particularly those who were tested more than 7 months after treatment initiation (indicated with * in Figure 2), potentially affecting the observed response rates.

3.3. Plasma AR status and disease progression

All 13 patients (100%) with AR alterations experienced tumor progression and demonstrated resistance to ongoing or prior ARSi therapy (Figures 3A–C and S1). In contrast, only 34.6% (9/26) of patients without detected AR alterations experienced disease progression while receiving ARSi treatment (Figures 3D and S2). Among the 13 patients with AR alterations, five initially presented a biochemical PSA response to ARSi therapy without detectable AR alterations during the response phase. However, AR alterations emerged subsequently, coinciding with radiographic or biochemical disease progression. Specifically, the detected AR alterations included two cases of AR amplifications, one patient with AR‐V7 expression, and two patients with hotspot mutations. At the time of AR alteration detection, the average PSA level was 4.38 μg/L (range: 1.41–12.3 μg/L) (Figures 3A,B and S1A–E). Notably, in one of these five cases (Figure S1B), an AR amplification was detectable prior to the rise in PSA levels, indicating that molecular progression may precede biochemical progression.

Following ARSi failure, six of the 13 AR alteration‐positive patients received chemotherapy, all of whom demonstrated improved responses compared to their prior ARSi treatment (Figures 3B and S1E–J). For four of these patients, liquid biopsy samples were available either before or after chemotherapy, while in two cases, samples were collected both before and after treatment. Notably, one patient exhibited a persistent hotspot mutation in the codon 702–742 region, detected both pre‐ and post‐chemotherapy, with only a slight decrease in the number of mutant copies per milliliter of plasma during treatment—from 40 copies/mL before therapy to 35 copies/mL after therapy (Figure 3B). Another patient had an AR amplification, with a modest change in copy number level, decreasing from 5.24 before chemotherapy to 3.65 after treatment (Figure S1H). These findings suggest that while chemotherapy can lead to clinical improvement in patients, AR alterations, such as mutations or amplifications, may persist despite treatment.

Additionally, three AR alteration‐positive patients who initially exhibited therapeutic response or stable disease later developed tumor progression with PSA levels exceeding 20 μg/L, coinciding with the emergence of AR alterations. However, liquid biopsy from the earlier response or stable disease phase (PSA levels below 20 μg/L) was unavailable for these patients. Among the three patients, two showed AR‐V7 expression alongside additional AR alterations, while one carried an AR amplification in combination with a mutation at codons 878–877 (Figures 3C and S1K–M). After approximately 20 months of ARSi therapy, PSA levels rose markedly in both AR‐V7 positive patients, with no response to treatment switch, indicating AR‐V7‐associated resistance. ¹³ , ¹⁵ The third patient progressed under abiraterone, likely due to the mutation at codons 878–877, which is known to expand AR ligand specificity and may be activated by abiraterone‐induced progesterone levels ¹² , ²¹ (Figures 3C and S1L).

In four cases with AR‐V7 positivity and/or AR amplification, longitudinal monitoring revealed dynamic changes in the concentration of AR‐V7 expression and AR CNA. Three patients showed rising AR CNAs alongside increasing PSA levels, with CN values shifting from 1.23, 1.63, and 1.97 to 1.35, 1.66, and 3.65, respectively (Figure S1K–L,H). One of them also developed AR‐V7 expression, which rose from 7.8 to 49 copies/mL under enzalutamide, indicating complete treatment failure (Figures 3C and S1K). The fourth patient, assessed post‐abiraterone, also showed a strong increase in PSA (304 to 1754 μg/L) and AR‐V7 (5.8 to 33.3 copies/mL; Figure S1G). These molecular alterations were accompanied by parallel upregulation of AR‐FL and KLK3 transcript levels, as illustrated in the heatmaps presented in Figure 2B,C.

Among the 26 AR alteration‐negative patients, nine experienced tumor progression. Of these, two exhibited primary resistance to ARSi therapy, showing no PSA decline after treatment initiation, while seven initially responded to ARSi but later developed tumor progression. Unfortunately, additional advanced stage information—such as markers for neuroendocrine differentiation, histopathological data, or genomic sequencing data—were unavailable for these patients, preventing conclusions regarding the mechanisms underlying their primary or secondary resistance to ARSi therapy (Figures 3D and S2). The remaining 17 AR alteration‐negative patients did not develop tumor progression during the study observation period and are depicted in Figure S3.

In summary, patients without AR alterations tend to have a more stable disease course, with relatively consistent PSA trajectories over time (Figure 3E). In contrast, AR‐alteration‐positive patients display more dynamic PSA fluctuations, characterized by both PSA increases and transient declines, indicative of a more heterogeneous treatment response and disease progression pattern (Figure 3F). Our findings further suggest that real‐time AR monitoring may detect resistance earlier than PSA alone. While AR‐V7 and AR CNA levels often correlate with PSA, AR alterations can support therapy decision‐making beyond conventional biomarkers.

3.4. Progression‐free survival and overall survival

Patients with AR alterations exhibited a significantly shorter median PFS of 11 months compared to 52 months in AR‐negative patients (p = 0.001; Figure 4A). Similarly, the median OS was shorter in patients with AR alteration (41 vs. 74 months) compared to those without detected AR alterations; however, this difference did not reach statistical significance (p = 0.0619; Figure 4B). Due to the limited number of progression and survival events, a stratified analysis based on the different categories of AR alterations was not feasible.

FIGURE 4.

Kaplan–Meier curves depicting (A) progression‐free survival and (B) overall survival stratified by AR alteration status (positive vs. negative). The number of patients at risk at each time point for both AR alteration‐positive and AR alteration‐negative groups is displayed below each plot. AR, androgen receptor.

To assess the prognostic significance of AR status, univariate Cox proportional hazard models were applied to evaluate time‐to‐event outcomes and test for independence from other clinical parameters. Among the variables analyzed, AR alterations emerged as an independent predictor for PFS (p = 0.0022), alongside anemia, while site score, PSA, and LDH were not statistically significant (Table 2). Due to the limited number of events in each cohort, multivariable models were not performed.

TABLE 2.

Univariate Cox proportional hazard analysis assessing independence from other clinical parameters.

Progression free survival			Overall survival
Variable	HR (95% CI)	p value	Variable	HR (95% CI)	p value
AR‐Status	3.992 (1.673 to 10.14)	0.0022	AR‐Status	2.618 (0.9368 to 7.864)	0.0699
Hb (>13, <13)	5.817 (2.028 to 19.03)	0.0016	Hb (>13, <13)	5.333 (1.781 to 15.66)	0.0021
Site Score (≤1, >1)	0.9722 (0.3867 to 2.280)	0.9496	Site Score (≤1, >1)	1.803 (0.5658 to 5.628)	0.3035
PSA μg/L	0.9985 (0.9928 to 1.002)	0.4967	PSA μg/L	0.9930 (0.9616 to 1.002)	0.3813
LDH (>250, <250 U/L)	1.965 (0.5471 to 5.673)	0.2443	LDH (>250, <250 U/L)	2.740 (0.7416 to 8.297)	0.0927

Abbreviations: AR‐Status, androgen receptor status; Hb (>13, <13), hemoglobin levels greater than 13 g/dL or less than 13 g/dL; HR (95% CI), hazard ratio (95% Confidence Interval); LDH (>250, <250 U/L), lactate dehydrogenase levels greater than 250 units per liter or less than 250 units per liter; PSA μg/L, prostate‐specific antigen concentration as a continuous variable in micrograms per liter; Site Score (≤1, >1), number of metastatic sites, with scores of 1 or less versus more than 1.

4. DISCUSSION

This observational study, conducted in a real‐world clinical setting, provides valuable insights into the clinical utility of AR alteration monitoring in mCRPC patients using liquid biopsy. Unlike controlled clinical trials, which often follow standardized protocols, this study reflects real‐world treatment dynamics, where therapeutic decisions are made based on the evolving clinical status of each patient. The high detection rate of AR alterations (33.3%) in this cohort, along with their association with inferior clinical outcomes ‐ such as lower PSA response rates and higher rates of disease progression ‐ supports the potential role of AR status assessment in routine clinical practice.

Patients harboring AR alterations exhibited poorer therapeutic responses, with a significantly lower PSA response rate to ARSi therapy (37.5%) compared to patients without AR alterations (80.7%). Given the distinct functional consequences of AR‐V7 expression, AR amplification, and hotspot mutations, we analyzed response rates for each alteration individually. ¹³ , ¹⁴ , ¹⁵ , ²² , ²³ , ²⁴ AR‐V7 was strongly associated with primary resistance, with no responders observed in this subgroup (0%) while AR amplification and AR hotspot mutations showed a trend toward reduced efficacy, albeit less distinctly (50% and 60%). These findings are consistent with data reported by Armstrong et al., showing a response rate of only 11% of AR‐V7–positive patients, compared to 30% in AR‐V7–negative cases, underscoring the limited response in this subgroup. ¹¹ Importantly, our findings demonstrate that not only AR‐V7 but also AR amplifications and point mutations negatively impact clinical response across all affected patients. Regardless of initial treatment response, all patients with AR alterations ultimately experienced disease progression, as indicated by increasing PSA levels at or shortly after the time of alteration detection. This observation is particularly relevant, as only a subset of our cohort was molecularly profiled before therapy initiation. Our data further show that AR alterations frequently co‐occur, with 54% of patients exhibiting multiple AR alterations, suggesting a cumulative effect on resistance. This is consistent with Ledet et al., who reported co‐existing AR amplification and/or hotspot mutations in approximately 40% of 892 mCRPC patients, ²⁵ and with Del Re et al., who identified AR amplification in combination with AR‐V7 expression in 9 of 43 positive patients, underscoring the frequent co‐occurrence of these resistance mechanisms. ¹⁵

Moreover, the median PFS was significantly shorter in AR alteration‐positive patients compared to negative patients (11 vs. 52 months, p = 0.001). Although the difference in OS (41 vs. 74 months) did not reach statistical significance (p = 0.0619), likely due to sample size limitations and treatment variability, the trend suggests that AR alterations may be associated with more aggressive disease phenotypes. Our results align with previous reports linking AR alterations to poor outcomes. Romanel et al. found that L702H and T878A mutations were significantly associated with shorter PFS and OS in abiraterone‐treated patients. ²⁶ Similarly, Conteduca et al. reported these mutations in 8 of 171 ARSi‐treated patients, also correlating with reduced OS. ¹⁶ Del Re et al. observed significantly worse median PFS in AR‐V7‐positive vs. negative patients (PFS: 5.4 vs. 24.3 months). ¹⁵

Our study underscores the importance of continuous AR monitoring throughout treatment rather than relying solely on baseline assessments. AR alterations were not always present at treatment initiation but emerged in some patients as the disease progressed. Specifically, five patients developed AR alterations after starting ARSi therapy, highlighting the dynamic nature of AR alterations during treatment. This finding aligns with previous research by Nakazawa et al., which demonstrated that AR‐V7 expression in circulating tumor cells can evolve and even disappear over the course of therapy. ²⁷

The emergence of AR alterations in patients who initially responded to ARSi but later developed resistance suggests a potential role in acquired treatment failure rather than primary resistance. Notably, in one longitudinally monitored case, an AR amplification was detected prior to a PSA rise, demonstrating that liquid biopsy can reveal molecular progression earlier than conventional markers and may enable more timely therapeutic interventions.

Our findings also suggest that patients with AR alterations may benefit more from taxane‐based chemotherapy than from ARSi therapy. Previous studies, including those by Antonarakis et al., have demonstrated that patients with AR‐V7‐positive metastatic CRPC respond more favorably to chemotherapy compared to ARSi. ⁸ , ¹¹ , ¹⁴ , ²⁸ Consistent with this, six patients in this study who transitioned to chemotherapy after failing ARSi therapy exhibited superior clinical responses. While this observation reinforces the potential predictive value of AR alterations in guiding treatment selection, it is based on a small sample size, and larger studies are needed to validate these findings. This is particularly relevant in light of a contrasting report by Palmberg et al., who described a patient with AR amplification showing a superior response to maximal androgen blockade compared to preceding chemotherapy. ²⁹ Nonetheless, these data contribute to the growing body of evidence suggesting that chemotherapy may be a more effective treatment option for AR alteration‐positive patients and highlight the potential role of AR biomarker testing in optimizing therapeutic sequencing.

This study has some limitations, including a small and heterogeneous sample size and the absence of standardized treatment protocols. However, these limitations also reflect the real‐world clinical scenario, where treatment decisions are often adapted to individual patient needs rather than following strict protocol‐driven interventions. Instead, this observational study provides pragmatic evidence of the potential relevance of AR alterations in the management of mCRPC in real‐world oncology practice.

In recent years, the therapeutic landscape of prostate cancer has evolved significantly, particularly with the earlier introduction of ARSi therapy in hormone‐sensitive stages of the disease. This shift in treatment sequencing has important implications for resistance mechanisms and therapeutic outcomes. Previous studies have reported limited efficacy of sequential ARSi use in advanced prostate cancer, emphasizing the need for biomarker‐driven treatment strategies to optimize patient selection and improve therapeutic outcomes. ⁶ , ³⁰ , ³¹ , ³² , ³³

Our study confirms the importance of longitudinal AR monitoring as an important additional tool in clinical practice. Liquid biopsy provides a minimally invasive, patient‐friendly, and effective method for serial monitoring of AR status. Unlike tissue biopsies, which are often challenging to obtain in advanced prostate cancer, liquid biopsy enables real‐time assessment of tumor evolution, allowing for earlier detection of treatment resistance and timely adjustments to therapeutic strategies. Moreover, the use of cfDNA and cfRNA in combination with ddPCR is cost‐effective and easily implementable in most laboratories, making it an accessible tool for widespread clinical application. This combination of affordability, accessibility, and real‐time monitoring capabilities highlights the clinical applicability of AR alteration screening as a biomarker‐driven tool in CRPC management.

In conclusion, this real‐world observational study demonstrates that monitoring AR alterations via liquid biopsy is a promising strategy for guiding treatment decisions in mCRPC. AR alterations are associated with reduced ARSi efficacy, shorter PFS, and a trend toward decreased OS. Therefore, patients with AR mutations may derive greater benefit from chemotherapy. The dynamic nature of AR alterations throughout disease progression underscores the necessity of continuous monitoring to optimize therapeutic strategies. However, future prospective studies with serial sampling before and during ARSi treatment are needed to better characterize the dynamics of AR alterations.

Integrating liquid biopsy‐based AR testing into routine clinical practice may improve personalized treatment approaches, enabling earlier detection of resistance and more informed therapeutic adjustments. These findings support the potential for AR biomarker testing to enhance patient outcomes in mCRPC, particularly as the field moves toward precision medicine in advanced prostate cancer care.

AUTHOR CONTRIBUTIONS

Regina Stitz: Conceptualization; writing – original draft; methodology; validation; visualization; project administration; data curation; formal analysis. Franz Stoiber: Investigation; writing – review and editing; project administration; data curation. Renè Silye: Supervision; resources; writing – review and editing; project administration. Elisabeth Rebhan: Data curation; methodology; writing – review and editing. Michael Dunzinger: Investigation; writing – review and editing; resources. Franz Pühringer: Writing – review and editing; investigation; methodology. Ellen Heitzer: Supervision; writing – original draft; writing – review and editing; methodology; formal analysis. Cornelia Hauser‐Kronberger: Writing – review and editing; project administration; supervision.

CONFLICT OF INTEREST STATEMENT

The authors declare no conflict of interest.

ETHICS STATEMENT

This study was conducted in accordance with the principles of the Declaration of Helsinki. Ethical approval was obtained from the Ethics Committee of Upper Austria (Approval No. 1271/2021). Written informed consent was obtained from all participants prior to their inclusion in the study. The Ethics Committee raised no objections to the conduct of the study.

Supporting information

FIGURE S1. Individual PSA progression in AR alteration positive patients. Panels A–M depict individual PSA trajectories (log₁₀ scale) for all AR alteration–positive patients, spanning from the initiation of first‐line ARSi therapy to either the end of study observation or patient death. Each panel includes the detection of AR status and highlights corresponding treatment interventions. Treatments are represented as ARSi therapy (light blue with various patterns), chemotherapy (pink), radioligand therapy (olive), or no systemic therapy (gray). AR, androgen receptor; SNVs*, AR p.T878A/S‐F877L + AR p.L702H‐W742L/C; ARSi, androgen signaling inhibitor therapy; ADT, androgen deprivation therapy; PSA, prostate‐specific antigen.

FIGURE S2. Individual PSA trajectories in AR alteration‐negative patients with biochemical, clinical, or radiographic progression. Panels A–I display individual PSA trajectories (log₁₀ scale) for nine AR alteration–negative patients who experienced tumor progression during the study period. Each trajectory spans from the initiation of first‐line ARSi therapy to either the end of study observation or the patient's death. Each panel includes the AR status and highlights relevant therapeutic interventions. Treatments are represented as ARSi therapy (light blue with various patterns), chemotherapy (pink), radioligand therapy (olive), or no systemic therapy (gray). Of the nine patients, two showed no PSA response immediately after the start of ARSi therapy (B and F), while seven initially responded to therapy but later developed tumor progression. AR, androgen receptor; ARSi, androgen receptor signaling inhibitor; ADT, androgen deprivation therapy; PSA, prostate‐specific antigen.

FIGURE S3. Individual PSA trajectories in AR alteration‐negative patients without biochemical, clinical, or radiographic progression. Panels A–Q display individual PSA trajectories (log₁₀ scale) of 17 patients who remained negative for AR alterations and exhibited no evidence of biochemical, clinical, or radiographic progression throughout the study period. Each trajectory spans from the initiation of first‐line ARSi therapy to either the end of study observation or the patient's death. Each panel includes the AR status and highlights relevant therapeutic interventions. Treatments are represented as ARSi therapy (light blue with various patterns), chemotherapy (pink), radioligand therapy (olive), or no systemic therapy (gray). All patients demonstrated durable treatment response and maintained stable disease throughout the observation period, with consistent AR alteration–negative status confirmed across timepoints. AR, androgen receptor; ARSi, androgen receptor signaling inhibitor; ADT, androgen deprivation therapy; PSA, prostate‐specific antigen.

Stitz R, Stoiber F, Silye R, et al. Clinical impact of real‐time androgen receptor alteration monitoring on metastatic castration‐resistant prostate cancer treatment in real‐world settings. Int J Cancer. 2025;157(10):2124‐2134. doi: 10.1002/ijc.70019

DATA AVAILABILITY STATEMENT

The data supporting the findings of this study are provided within the article and its supplementary material. Further information is available from the corresponding author upon request.