Phase Ib Trial of Bavdegalutamide (ARV-110) in Combination with Abiraterone for Metastatic Prostate Cancer

Abstract

Purpose:

Resistance to androgen receptor (AR) pathway inhibitors necessitates the development of regimens to improve outcomes in metastatic prostate cancer (mPC). This phase Ib clinical trial evaluated the safety/tolerability and pharmacokinetics of bavdegalutamide, an oral proteolysis-targeting chimera AR degrader, combined with abiraterone in patients with mPC experiencing prostate-specific antigen (PSA) progression on abiraterone.

Patients and Methods:

Patients received bavdegalutamide 420 mg, abiraterone ≤1,000 mg, and corticosteroid daily. Assessments included safety [dose-limiting toxicities (DLT); treatment-emergent adverse events (TEAE)], pharmacokinetics, PSA control rate, ≥50% PSA decline (PSA₅₀), and radiographic progression-free survival (rPFS).

Results:

Forty-five patients were treated; seven had AR ligand-binding domain (LBD) mutations. The median duration of exposure was 35.7 weeks; 13 patients were continuing bavdegalutamide at the data cutoff. No DLTs were observed. There were no clinically significant effects of bavdegalutamide on abiraterone pharmacokinetics. Bavdegalutamide 420 mg daily was selected as the recommended phase II dose. TEAEs were mostly grade 1/2 (82.2%), and 17.8% were grade 3. Dose modifications or discontinuations due to TEAEs occurred in 48.9% and 13.3% of patients, respectively. The most common bavdegalutamide-related TEAEs were fatigue (48.9%) and nausea (46.7%). The PSA control rate was 48.9% [95% confidence interval (CI), 33.7–64.2]. The PSA₅₀ response rate was 24.4% (95% CI, 12.9–39.5), including in five out of seven patients with AR LBD mutations. The median time to PSA progression was 2.8 months (95% CI, 1.5–6.5). The median rPFS was 16.3 months (95% CI, 8.2–not calculable).

Conclusions:

Bavdegalutamide combined with abiraterone demonstrated a manageable safety profile with no major drug–drug interactions and evidence of clinical activity in mPC.

Translational Relevance.

This phase Ib clinical trial evaluated the safety, tolerability, and pharmacokinetics of the proteolysis-targeting chimera (PROTAC) androgen receptor (AR) degrader bavdegalutamide in combination with abiraterone in patients with metastatic prostate cancer. The study advances our understanding of combinatorial strategies targeting the AR pathway, addressing a critical unmet need for patients who develop resistance to current AR pathway inhibitors (ARPI). In this study, bavdegalutamide in combination with abiraterone had a manageable safety profile. No major drug–drug interactions between abiraterone and bavdegalutamide occurred, and no effects on abiraterone pharmacokinetic parameters were observed. Additionally, the combination regimen demonstrated promising antitumor activity. These data support further development of a PROTAC AR degrader in combination with ARPIs for the treatment of prostate cancer.

Introduction

Globally, prostate cancer is the second most common malignancy in men, accounting for approximately 14.2% of all cancer cases and 7.3% of male cancer deaths worldwide (1). In the United States, prostate cancer is the most frequently diagnosed cancer among men, with more than 2.1 million cases recorded between 2012 and 2021 and an incidence rate of approximately 286.9 per 100,000 (2). Despite advances in early detection and localized treatment, the incidence of metastatic prostate cancer (mPC) has increased across all age and racial groups over the past decade in the United States (3). The 5-year survival rate for mPC in the United States is 38% (3).

The androgen receptor (AR) pathway is critical for the initiation and progression of prostate cancer and thus represents a key therapeutic target (4, 5). Current US FDA-approved second-generation AR pathway inhibitors (ARPI) for patients with metastatic hormone-sensitive prostate cancer (mHSPC) and metastatic castration-resistant prostate cancer (mCRPC) include enzalutamide, apalutamide, darolutamide, and abiraterone. The oral small-molecule ARPI abiraterone blocks intratumoral and adrenal androgen biosynthesis (5–9) and is approved in combination with prednisone and androgen deprivation therapy (ADT) for patients with mCRPC or high-risk mHSPC (10). A key limitation of abiraterone, and ARPI therapy in general, is that many patients with mCRPC will have resistance due to prior ARPI treatment, and nearly all patients who initially respond to treatment will eventually develop resistance, leading to disease progression (7, 8, 11, 12). The majority of resistant tumors maintain reliance on AR signaling through mechanisms that include increased AR expression and AR mutations occurring most often in the ligand-binding domain (LBD; amino acids 671–920; refs. 9, 13–15). In patients with mCRPC, AR LBD mutations have been observed in 15% of ARPI-naïve patients, with prevalence increasing in the setting of ARPI resistance, occurring in 22% and 24% of patients with mCRPC after one or two lines of second-generation ARPI therapy, respectively (15). The most common AR LBD mutations after abiraterone treatment are L702H, T878A/S, and H875Y (15). AR LBD mutations affect treatment response; for example, patients with T878A or L702H point mutations or AR gain before abiraterone treatment were 4.9 times less likely to have a ≥50% decline in prostate-specific antigen (PSA₅₀) than patients with normal plasma AR prior to treatment (16). Furthermore, patients with prostate cancer with AR LBD mutations have reduced overall survival in real-world settings (15, 17–19). Therefore, the development of novel therapeutic strategies targeting AR for the treatment of mCRPC and mHSPC is a critical unmet need for this patient population.

Bavdegalutamide (ARV-110) is a novel, oral proteolysis-targeting chimera (PROTAC) that harnesses the ubiquitin–proteasome system to degrade AR (20, 21, 22). Bavdegalutamide targets wild-type AR and clinically relevant AR LBD-mutant proteins (22, 23). In preclinical in vitro models, bavdegalutamide degraded all tested AR LBD-mutant proteins associated with abiraterone and/or enzalutamide resistance although AR L702H was degraded with lesser potency (22). Furthermore, the benefits of the degradation-based mechanism of action were highlighted by the superior activity of bavdegalutamide compared with the ARPI enzalutamide in cellular models evaluating PSA production, prostate cancer cell proliferation, and apoptosis induction. Bavdegalutamide also demonstrated in vivo tumor growth inhibition in noncastrated and castrated mouse xenograft models, including those resistant to enzalutamide and abiraterone, supporting its evaluation in patients with mCRPC who develop ARPI resistance (22).

Bavdegalutamide monotherapy was evaluated in a phase I/II study (NCT03888612) in patients with mCRPC after up to two prior hormonal agents (e.g., abiraterone and/or enzalutamide), including patients who had received prior chemotherapy, with 420 mg once daily selected as the recommended phase II dose (RP2D; ref. 24). Bavdegalutamide demonstrated promising single-agent antitumor activity in patients pretreated with ARPIs, with enhanced activity observed in a subgroup of patients with tumors harboring AR T878X and/or H875Y missense LBD mutations (24).

Based on complementary mechanistic and pharmacokinetic activity to bavdegalutamide, abiraterone was chosen over AR antagonists (e.g., enzalutamide, which, like bavdegalutamide, also targets the AR LBD) as a preferred combination partner to bavdegalutamide. Although prior studies evaluating combinations of abiraterone with other next-generation ARPIs, such as enzalutamide and apalutamide, have not shown promise (25–27), bavdegalutamide’s potential to more fully block AR signaling via protein degradation warranted investigation in combination with abiraterone by targeting multiple mechanisms of ARPI resistance. The combination of bavdegalutamide and abiraterone was initially tested preclinically in a prostate tumor xenograft model and demonstrated combinatorial activity, highlighting the clinical potential for this dual AR pathway inhibition (22).

To assess the clinical utility of this combinatorial approach, we conducted this phase Ib study to determine if the addition of bavdegalutamide to an abiraterone-based regimen at the initiation of biochemical progression on abiraterone (PSA progression without radiographic progression) may overcome abiraterone resistance and reestablish the efficacy of AR pathway blockade in patients with mPC.

Patients and Methods

Study design

This was an open-label, multicenter, nonrandomized phase Ib study evaluating the safety, tolerability, pharmacokinetics, and preliminary antitumor activity of bavdegalutamide in combination with abiraterone and a concomitant corticosteroid in adult male patients with mPC and increasing PSA on abiraterone therapy (ClinicalTrials.gov Identifier: NCT05177042). The study was conducted between February 1, 2022, and April 10, 2025, at 18 sites across Canada, France, the United Kingdom, and the United States. A safety lead-in cohort of six patients was conducted, and a safety review committee evaluated all available safety and pharmacokinetic data to determine the RP2D before additional patients were treated (Supplementary Fig. S1A).

The study was conducted according to the protocol and consensus ethical principles derived from the Declaration of Helsinki, the International Ethical Guidelines for Biomedical Research Involving Human Subjects, the Belmont Report, and the US Common Rule (45 CFR 46). The protocol was approved by Institutional Review Boards or independent ethics committees before study initiation. Written informed consent was obtained from all participants prior to the performance of any protocol-related procedures.

Patients

Full eligibility criteria are available in Supplementary Appendix S1. Eligible patients were adults who had mCRPC or mHSPC and were receiving ongoing treatment with stable doses of abiraterone and a concomitant corticosteroid. Patients were to have increasing PSA values, occurring at least 16 weeks after the initiation of abiraterone, with at least two increasing PSA values higher than the PSA nadir on abiraterone, and no known radiographic evidence of disease progression. Patients were to receive ongoing ADT with gonadotropin-releasing hormone (GnRH) analogs/inhibitors or have received a prior orchiectomy and had to have an Eastern Cooperative Oncology Group performance status of 0 or 1. Key exclusion criteria included previous treatment with enzalutamide, apalutamide, darolutamide, or experimental AR-directed therapies; receipt of chemotherapy, investigational agents, immunotherapy, or hormonal therapy other than GnRH agonists within 28 days of the start of study treatment; or radiotherapy within 4 weeks of the start of study treatment or previous irradiation to greater than 25% of the bone marrow.

Interventions

In this single-arm study, there was no blinding of participants or investigators, and all patients received oral bavdegalutamide 420 mg daily, abiraterone up to 1,000 mg daily (i.e., the same dose as before study enrollment), and a concomitant corticosteroid (dosed at the investigator’s discretion at the same dose as before study entry; dose modifications or switching of corticosteroids were to be reported to the study sponsor) on 28-day cycles.

Assessments

Safety assessments

The primary objectives were to evaluate the safety and tolerability of bavdegalutamide in combination with abiraterone and to select the RP2D and schedule for the expansion phase of the trial. The safety endpoints included dose-limiting toxicities (DLT) in the first 4 weeks of treatment and RP2D determination based on safety and available pharmacokinetic data from the safety lead-in (criteria for the determination of DLTs are summarized in Supplementary Appendix S2). Based on Bayesian optimal interval design modeling (28), enrollment continued without dose de-escalation if the DLT rate was below 0.358.

The frequency, seriousness, severity, and relationship to the study drug of adverse events (AE) and laboratory abnormalities (severity categorized according to NCI Common Terminology Criteria for Adverse Events version 5.0) were also assessed. AEs were coded using the Medical Dictionary for Regulatory Activities (version 20.1 or higher).

Pharmacokinetic assessments

The pharmacokinetic parameters of abiraterone given alone or in combination with bavdegalutamide were assessed as a secondary outcome using plasma concentration–time data and included the determination of the area under the plasma concentration–time curve to the end of the dosing period (AUC_0–tau), area under the plasma concentration–time curve from the time of dosing to the last measurable concentration (AUC_last), maximum plasma concentration (C_max), minimum plasma concentration (C_min), last measurable plasma concentration (C_last), time of maximum plasma concentration (T_max), and time of last plasma concentration (T_last).

Pharmacokinetic samples were collected from patients in the safety lead-in before and 1, 2, 4, 6, and 8 hours after abiraterone dosing on day −1 and at the same time points on cycle 1, day 21, as well as before and 1 hour after abiraterone administration on cycle 2, day 1. Pharmacokinetic parameters for abiraterone alone are reported for day −1, and pharmacokinetic parameters for abiraterone and bavdegalutamide are reported for cycle 1, day 21. On day −1, abiraterone was administered on an empty stomach (i.e., ≥1 hour before or ≥2 hours after a meal). On days 1 and 21 of cycle 1 and day 1 of all subsequent cycles, abiraterone was administered on an empty stomach, and then a sample was collected for the assessment of pharmacokinetics ≥1 hour after the abiraterone dose. Once the sample was taken, bavdegalutamide was administered with a meal. The 1-hour postdose sample on these days was taken before bavdegalutamide administration.

Efficacy assessments

Antitumor activity was assessed as a secondary outcome. Efficacy assessments included the PSA response rate, which was defined as the percentage of participants with a ≥30% or ≥50% decrease in PSA values from baseline (PSA₃₀ and PSA₅₀, respectively) that was confirmed to be sustained by a second PSA value obtained at least 3 weeks later. Additionally, the percentage of patients with PSA control (i.e., lack of PSA progression at 12 weeks, in which PSA progression was defined as at least a 25% increase in PSA and an absolute increase in PSA of at least 2 ng/mL above the nadir, which was confirmed by a second consecutive value obtained at least 3 weeks later) and time to PSA progression (i.e., time interval from the date of the first bavdegalutamide dose to the date of PSA progression, with participants without progressing PSA censored on the date of the last PSA assessment before receipt of new anticancer therapy) were determined. PSA-related endpoints were defined using Prostate Cancer Working Group 3 (PCWG3) criteria. Additionally, the best percentage change in PSA level from baseline (i.e., maximum decrease or minimum increase in each patient) was determined.

A confirmed objective response rate (ORR; i.e., confirmed complete response or partial response) was determined for patients with measurable disease at baseline, using modified Response Evaluation Criteria in Solid Tumors (RECIST) 1.1 and PCWG3 criteria from computed tomography, magnetic resonance imaging, and bone scans performed every 8 weeks for the first 6 months of study treatment, then every 12 weeks until the third year of study treatment, and every 6 months thereafter on target, nontarget, and new lesions.

Radiographic progression-free survival (rPFS; i.e., time interval from the date of the first bavdegalutamide dose to the date of progression per modified RECIST 1.1/PCWG3 or death due to any cause) was assessed.

Post hoc efficacy analyses

Additional post hoc analyses were conducted in a subgroup of patients, which included those with mCRPC at abiraterone start (group 1) or those with mHSPC at abiraterone start with evidence of confirmed modified PCWG3 PSA progression (group 2). These analyses included PSA responses, time-to-event outcomes, and treatment duration, with subgroup evaluations (e.g., AR LBD mutation status, prior docetaxel, and corticosteroid use) as outlined in Supplementary Appendix S3.

Biomarker analyses

The mutation and amplification status of AR was assessed via circulating tumor DNA (ctDNA) testing at a central laboratory using the FoundationOne Liquid CDx assay (Foundation Medicine, Inc.) as described previously (29).

Statistical analyses

No formal hypothesis testing was planned for this phase Ib study. Therefore, no formal sample size or power calculation was performed. Approximately 40 patients were to be treated to evaluate the safety profile and preliminary antitumor activity of bavdegalutamide in combination with abiraterone. Assuming the target PSA control rate (as measured by lack of PSA progression at 12 weeks) was between 30% and 90%, the width of the corresponding 95% confidence intervals (CI) per the Clopper–Pearson exact method would be 0.21 to 0.33. The study populations included for the various assessments are summarized in Supplementary Table S1.

Descriptive statistics for pharmacokinetic parameters [i.e., number, mean and/or median, standard deviation (SD), coefficient of variation, and range] were tabulated by cohort, dose level, analyte, cycle, and day in the pharmacokinetic data analysis set. A comparison of the natural log-transformed pharmacokinetic parameters C_max, AUC_last, and AUC_tau of abiraterone was made to evaluate the relative bioavailability of abiraterone and bavdegalutamide (test) versus abiraterone alone, with data captured before study drugs were administered (reference). A mixed linear model, which included treatment as fixed effects and patients as a random effect, was performed. Least square means (LSM), the difference between LSMs, and 90% CIs of the difference were exponentiated to the original scale. Geometric LSMs, percentage of geometric mean ratios, and 90% CIs were tabulated.

Analysis of binary endpoints such as PSA control rate, PSA response rate, and ORR was based on all corresponding response-evaluable patients. The estimates of these parameters and the corresponding two-sided 95% CI were calculated using the Clopper–Pearson exact method.

Time-to-event endpoints such as time to PSA progression were estimated using the Kaplan–Meier product-limit method.

Additional post hoc analyses were performed using descriptive statistics and time-to-event methods as specified in Supplementary Appendix S3.

Results

Patients and treatment

Overall, 45 patients were treated (Supplementary Fig. S1B), including six patients in the safety lead-in. The median patient age was 68 years (range, 53–85; Table 1; Supplementary Table S2). All patients were receiving abiraterone 1,000 mg daily before study enrollment [median duration of prior abiraterone treatment, 96.4 weeks (range, 5–425)], including 34 (75.6%) patients for castration-resistant disease and 11 (24.4%) for hormone-sensitive disease at the time of abiraterone initiation. In all, 14 (31.1%) patients had received prior chemotherapy, and 44 (97.8%) were receiving corticosteroids at baseline (prednisolone, n = 35; dexamethasone, n = 9; unknown, n = 1). The median PSA at baseline was 4.99 ng/mL (range, 0.3–86.3).

Table 1.

Baseline patient and disease characteristics.

Characteristic	All patients (N = 45)
Age, years	​
Median (range)	68 (53–85)
Race, n (%)	​
White	21 (46.7)
Black	4 (8.9)
Missinga	20 (44.4)
Baseline ECOG performance status, n (%)	​
0	37 (82.2)
1	8 (17.8)
PSA, ng/mL	​
Median (range)	4.99 (0.3, 86.3)
​	(N = 43)
AR LBD mutation statusb, n (%)	7 (16.3)
Disease status at start of abiraterone prior to study entry, n (%)
Castration-resistant	34 (75.6)
Hormone-sensitive	11 (24.4)
Duration of abiraterone treatment prior to study entry, weeks	​
Mean (SD)	120.3 (78.7)
Median (range)	96.4 (5–425)
Prior systemic anticancer therapy, n (%)	​
GnRH analog or inhibitor	45 (100)
Abiraterone	45 (100)
First-generation antiandrogen	16 (35.6)
Chemotherapy	14 (31.1)
Bone metastases treatment	3 (6.7)
Targeted therapy	1 (2.2)
Other agents	44 (97.8)
Corticosteroid use at baseline, n (%)	​
Prednisolone	35 (77.8)
Dexamethasone	9 (20)
Unknown	1 (2.2)

Data are for the all-treated analysis set.

Abbreviation: ECOG, Eastern Cooperative Oncology Group.

^aData were not collected at all study sites due to variation in reporting requirements by region.

^bForty-three patients had samples suitable for mutational analysis and detectable ctDNA at baseline.

In total, 43 of 45 patients had samples suitable for mutational analysis, all of whom had detectable ctDNA at baseline. Of these, seven (16.3%) patients had AR LBD mutations [five patients had T878/H875 mutations (one patient had co-occurring V716M, R630Q); two patients had mutations in 671 to 920 (other than T878/H875; specifically, Q799E, n = 1; W742C, W742L, n = 1); no patients had L702H mutations (Table 1)]. AR genotype and AR-V7 circulating tumor cell positivity are shown by individual patient in Supplementary Table S3 (AR-V7 status determination is described in Supplementary Appendix S4).

At the time of data cutoff (September 10, 2024), 13 patients were continuing to receive bavdegalutamide treatment, and 32 patients had discontinued treatment, with the most common reason for treatment discontinuation being clinical progression (Supplementary Fig. S1B).

Overall, the median duration of exposure and mean relative dose intensity for bavdegalutamide was 35.7 weeks (range, 2.3–94.1) and 92.1% (SD, 10.5%), respectively. Corresponding values for abiraterone were 35.7 weeks (range, 2.3–94.1) and 104.3% (SD, 26.6%). Although corticosteroid choice was at the discretion of the investigator, no patients switched corticosteroids during the study. In the 13 patients who remained on treatment at data cutoff (Fig. 1), the median duration of bavdegalutamide therapy was 66 weeks (range, 51–94.1). Of the patients remaining on treatment, two had AR LBD mutations, four patients had received prior docetaxel treatment (one patient had both prior docetaxel treatment and AR LBD mutations), and four patients were receiving dexamethasone (Fig. 1).

Figure 1.

Swimmer plot of duration of exposure to bavdegalutamide. Duration of exposure in weeks as of September 10, 2024, was calculated as [(last dose date – first dose date) + 1]/7. The post hoc efficacy cohort was defined as either patients with mCRPC at abiraterone start (group 1) or patients with mHSPC at abiraterone start with PSA progression (group 2).

Safety

No DLTs were observed during the first treatment cycle within the safety lead-in. All patients experienced at least one treatment-emergent AE (TEAE), which were grade 1 to 2 in 82.2% of 45 treated patients (Table 2). Grade 3 TEAEs occurred in 17.8% of patients, and no grade 4 or 5 TEAEs were reported. The most common TEAEs were fatigue (57.8%; all grade 1–2) and nausea (53.3%; all grade 1–2).

Table 2.

TEAEs occurring in ≥10% of patients.

Preferred term	Bavdegalutamide + abiraterone (N = 45)
	Grade 1 n (%)	Grade 2 n (%)	Grade 3 n (%)	Grades 4–5 n (%)	Any grade n (%)
All-causality TEAEs
Any	6 (13.3)	31 (68.9)	8 (17.8)a	0	45 (100)
Fatigue	17 (37.8)	9 (20)	0	0	26 (57.8)
Nausea	18 (40)	6 (13.3)	0	0	24 (53.3)
Alopecia	10 (22.2)	7 (15.6)	0	0	17 (37.8)
Decreased appetite	13 (28.9)	4 (8.9)	0	0	17 (37.8)
Diarrhea	12 (26.7)	3 (6.7)	0	0	15 (33.3)
Vomiting	12 (26.7)	1 (2.2)	0	0	13 (28.9)
Dry skin	6 (13.3)	5 (11.1)	0	0	11 (24.4)
Arthralgia	8 (17.8)	3 (6.7)	0	0	11 (24.4)
Constipation	8 (17.8)	2 (4.4)	0	0	10 (22.2)
Headache	7 (15.6)	2 (4.4)	0	0	9 (20)
Dysgeusia	9 (20)	0	0	0	9 (20)
Anemia	7 (15.6)	1 (2.2)	0	0	8 (17.8)
Lymphocyte count decreased	0	5 (11.1)	1 (2.2)	0	6 (13.3)
Insomnia	4 (8.9)	2 (4.4)	0	0	6 (13.3)
Cough	5 (11.1)	1 (2.2)	0	0	6 (13.3)
Dyspnea	6 (13.3)	0	0	0	6 (13.3)
COVID-19	3 (6.7)	2 (4.4)	0	0	5 (11.1)
Weight decreased	3 (6.7)	2 (4.4)	0	0	5 (11.1)
Fall	3 (6.7)	2 (4.4)	0	0	5 (11.1)
Edema peripheral	5 (11.1)	0	0	0	5 (11.1)
Bavdegalutamide treatment-related TEAEs
Any	14 (31.1)	24 (53.3)	3 (6.7)b	0	41 (91.1)
Fatigue	14 (31.1)	8 (17.8)	0	0	22 (48.9)
Nausea	15 (33.3)	6 (13.3)	0	0	21 (46.7)
Alopecia	9 (20)	7 (15.6)	0	0	16 (35.6)
Decreased appetite	10 (22.2)	4 (8.9)	0	0	14 (31.1)
Diarrhea	10 (22.2)	1 (2.2)	0	0	11 (24.4)
Vomiting	8 (17.8)	1 (2.2)	0	0	9 (20)
Dry skin	3 (6.7)	5 (11.1)	0	0	8 (17.8)
Dysgeusia	8 (17.8)	0	0	0	8 (17.8)
Headache	5 (11.1)	2 (4.4)	0	0	7 (15.6)
Weight decreased	3 (6.7)	2 (4.4)	0	0	5 (11.1)

^aAdditional grade 3 TEAEs not shown by preferred term were upper gastrointestinal hemorrhage, seizure, spinal cord compression, COVID-19 pneumonia, lymphocyte count decreased, platelet count decreased, white blood cell count decreased, and febrile neutropenia (each reported in one patient, 2.2%).

^bGrade 3 TEAEs considered related to bavdegalutamide and not shown by preferred term were platelet count decreased, lymphocyte count decreased, and white blood cell count decreased (each reported in one patient, 2.2%).

Overall, at least one dose adjustment was required for bavdegalutamide in 22 (48.9%) patients and for abiraterone in 15 (33.3%) patients because of TEAEs (Supplementary Table S4). Additionally, six (13.3%) patients experienced TEAEs leading to discontinuation of bavdegalutamide, and five (11.1%) patients experienced TEAEs leading to discontinuation of abiraterone.

Overall, 41 (91.1%) patients experienced TEAEs considered related to bavdegalutamide, and 20 (44.4%) patients experienced TEAEs considered related to abiraterone; three (6.7%) patients and one (2.2%) patient experienced grade 3 TEAEs that were considered related to bavdegalutamide and abiraterone treatment, respectively (Supplementary Table S4). The most common bavdegalutamide-related AEs were fatigue (48.9%; all grade 1–2) and nausea (46.7%; all grade 1–2; Table 2). Three (6.7%) patients experienced grade 3 bavdegalutamide-related AEs [lymphocyte count decreased, platelet count decreased, and white blood cell count decreased, each reported in one (2.2%) patient; Table 2], with the event of white blood cell count decreased considered related to both bavdegalutamide and abiraterone. No clinical sequelae were associated with these laboratory abnormalities.

Serious TEAEs were reported in four (8.9%) patients [one event each of febrile neutropenia (occurring from days 265–271), upper gastrointestinal hemorrhage (occurring from days 85–91 and days 309–323), COVID-19 pneumonia (occurring from days 51–88), and spinal cord compression (occurring from days 119–124)]; all were grade 3, and none were considered related to study treatment.

Pharmacokinetics

Pharmacokinetic parameters for abiraterone alone and when administered with bavdegalutamide were assessed in the six patients included in the safety lead-in. Changes in pharmacokinetic parameters for abiraterone alone (as measured on day –1) compared with those when abiraterone was administered with bavdegalutamide (as measured on cycle 1, day 21), including C_max, AUC_last, and AUC_tau, were not clinically significant (Table 3; Fig. 2; Supplementary Table S5). Based on these pharmacokinetic data and the safety data, the RP2D was confirmed to be 420 mg bavdegalutamide daily, administered with abiraterone and prednisone in a 28-day cycle.

Table 3.

Pharmacokinetic parameters.

Parameter	Abiraterone 1,000 mg Day –1 (n = 6)	Abiraterone 1,000 mg + bavdegalutamide 420 mg Cycle 1, day 21 (n = 6)
	Abiraterone	Abiraterone	Bavdegalutamide
Cmax (ng/mL)	​	​	​
Geometric mean (CV%)	88.20 (80.36)	108.96 (43.10)	423.1 (39)
Cmin (ng/mL)	​	​	​
Geometric mean (CV%)	12.72 (60.77)	6.94 (57.92)	293.6 (42.1)
Clast (ng/mL)	​	​	​
Geometric mean (CV%)	20.98 (55.34)	16.62 (29.98)	376.2 (31.9)
Tmax (hours)	​	​	​
Median (range)	2 (2–2)	1.84 (0.93–2.35)	4.9 (0–7)
Tlast (hours)	​	​	​
Median (range)	8 (8–8)	7.88 (7.60–7.93)	6.8 (7.7–7)
AUCtau (hours·ng/mL)	​	​	​
Geometric mean (CV%)	476.14 (56.55)	396.85 (33.73)	8,629.88 (35.61)
AUClast (hours·ng/mL)	​	​	​
Geometric mean (CV%)	348.71 (52.44)	317.98 (37.51)	2,393.87 (38.63)

Data are for patients in the safety lead-in and for the pharmacokinetic analysis set. Treatment columns are based on the initial dose group, defined as the cycle 1, day 1 dose/frequency received before pharmacokinetic samples were obtained.

Abbreviations: AUC_last, area under the plasma concentration–time curve from the time of dosing to the last measurable concentration; AUC_tau, area under the plasma concentration–time curve to the end of the dosing period; C_last, last measurable plasma concentration; C_max, maximum plasma concentration; C_min, minimum plasma concentration; CV, coefficient of variation; T_last, time of last plasma concentration; T_max, time of maximum plasma concentration.

Figure 2.

Ladder plots for pairwise comparisons of (A) AUC_last, (B) AUC_tau, and (C) C_max pharmacokinetic parameters following administration of abiraterone alone or with bavdegalutamide. Data are for patients in the safety lead-in and for the pharmacokinetic analysis set. Treatment groups are based on the initial dose group, defined as the cycle 1, day 1 dose/frequency received before pharmacokinetic samples were obtained. For abiraterone alone, parameters were measured on day –1 (–7 days). For abiraterone administered with bavdegalutamide, parameters were measured on cycle 1, day 21. AUC_0–tau, area under the plasma concentration–time curve to the end of the dosing period; AUC_last, area under the plasma concentration–time curve from the time of dosing to the last measurable concentration; C_max, maximum plasma concentration.

Efficacy

All patients

All 45 patients were included in the PSA efficacy analysis set. The PSA control rate was 48.9% (95% CI, 33.7–64.2). The rates of confirmed PSA decline of ≥30% and ≥50% (i.e., PSA₃₀ and PSA₅₀) were both 24.4% (95% CI, 12.9–39.5) overall (Fig. 3); in seven patients who had AR LBD mutations, four had confirmed PSA₃₀ and PSA₅₀ responses, three of whom had AR T878/H875 mutations. In the overall population, seven (15.6%) patients had achieved PSA <0.2 ng/mL in the postbaseline period. The median time to PSA progression among the overall population was 2.8 months (95% CI, 1.5–6.5; Supplementary Fig. S2A).

Figure 3.

Waterfall plot of the best percentage change in PSA value from baseline. Data are for the efficacy analysis set. The two-sided 95% CI is based on the Clopper–Pearson exact method. The post hoc efficacy cohort was defined as either patients with mCRPC at abiraterone start (group 1) or patients with mHSPC at abiraterone start with PSA progression (group 2).

Of 42 response-evaluable patients, 4.8% achieved a partial response, 78.6% had stable disease, 7.1% had progressive disease, and 9.5% had no evidence of disease at the data cutoff. Of the five patients with measurable disease at baseline, two achieved a partial response.

Overall, a total of 14 (31.1%) of 45 patients had an rPFS event, and the median rPFS was 16.3 months (95% CI, 8.2–not calculable; Supplementary Fig. S3A).

Post hoc efficacy analyses

Forty-two patients were included in the post hoc efficacy analyses, which included 34 patients with mCRPC at abiraterone start (group 1) and eight patients with mHSPC at abiraterone start with PSA progression (group 2). Three patients who had mHSPC at abiraterone start were excluded from the post hoc analyses because they did not have confirmed PSA progression by PCWG3 criteria at the time of bavdegalutamide initiation.

The median duration of treatment in the post hoc efficacy cohort was 31.7 weeks (range, 2.3–94.1) in the 42 patients included in the post hoc analyses and was generally consistent according to AR LBD mutation status [positive vs. negative: 36.8 (range, 4–87.9) weeks vs. 27.4 (range, 2.3–94.1) weeks] and receipt of prior docetaxel therapy [yes vs. no: 35.7 (range, 2.3–75) weeks vs. 27.6 (range, 7.9–94.1) weeks (Supplementary Table S6; Supplementary Fig. S4). At data cutoff, 12 patients included in the post hoc analyses remained on treatment, including two patients with AR LBD mutations, four patients with prior docetaxel treatment (one patient had both prior docetaxel treatment and AR LBD mutations), and four patients receiving dexamethasone (Fig. 1).

In the 34 patients with mCRPC at abiraterone start (group 1), the PSA control rate in the post hoc analysis was 47.6% (95% CI, 32–63.6; Supplementary Table S6; Supplementary Fig. S4A). Of the six patients harboring AR LBD mutations, four patients (66.7%) achieved PSA control (95% CI, 22.3–95.7) compared with 44.4% (16/36; 95% CI, 27.9–61.9) of patients not harboring AR LBD mutations. When evaluated by prior docetaxel use, 38.5% (5/13; 95% CI, 13.9–68.4) and 51.7% (15/29; 95% CI, 32.5–70.6) of patients with and without prior docetaxel use, respectively, achieved PSA control (Supplementary Table S6).

Confirmed PSA₅₀ and PSA₉₀ responses in the 42 patients in the post hoc efficacy group were seen in 10 (23.8%; 95% CI, 12.1–39.5) and 7 (16.7%; 95% CI, 7–31.4) patients, respectively (Supplementary Table S6; Supplementary Fig. S4B). Of the six patients harboring AR LBD mutations, four (66.7%; 95% CI, 22.3–95.7) achieved PSA₅₀, and three (50%; 95% CI, 11.8–88.2) achieved PSA₉₀. Corresponding percentages for the 36 patients without AR LBD mutations were 16.7% (95% CI, 6.4–32.8) and 11.1% (95% CI, 3.1–26.1). Of the 13 patients with prior docetaxel use, four (30.8%; 95% CI, 9.1–61.4) achieved PSA₅₀, and three (23.1%; 95% CI, 5–53.8) achieved PSA₉₀ compared with 20.7% (95% CI, 8–39.7) and 13.8% (95% CI, 3.9–31.7) of 29 patients without prior docetaxel use who achieved PSA₅₀ and PSA₉₀, respectively (Supplementary Table S6; Supplementary Fig. S4B). The median time to PSA progression in the post hoc efficacy group is shown in Supplementary Fig. S2B.

Fourteen (33.3%) of the 42 patients included in the post hoc analyses had an rPFS event, and the median rPFS was 16.3 months (95% CI, 8.2–not calculable; Supplementary Fig. S3B). In the 34 patients with mCRPC at abiraterone start (group 1), the median rPFS was 11.1 months (95% CI, 8.2–not calculable). In the eight patients with mHSPC at abiraterone start with PSA progression (group 2), the median rPFS was 16.3 months (95% CI, 3.5–not calculable; Supplementary Fig. S3B).

Discussion

mCRPC is characterized by AR signaling pathway reactivation despite ADT, as well as overexpression and mutations in AR (30–32). Additionally, a subset of patients with mCRPC have AR-mediated primary resistance to abiraterone, and most experience secondary resistance and disease progression during treatment (7, 8), emphasizing the urgent need for new therapeutic approaches to improve outcomes for these patients.

This phase Ib study reports the first early-phase clinical data for the combination of the PROTAC AR degrader bavdegalutamide with a standard-of-care ARPI. Bavdegalutamide in combination with abiraterone had a manageable safety profile in patients with mPC, with a low rate of high-grade TEAEs that was largely consistent with the toxicity profiles of these agents when administered as monotherapy (24, 33). Abiraterone dosing frequency should be increased if strong concomitant inhibitors of cytochrome P450 3A4 (CYP3A4) are coadministered (10), and in vitro data demonstrated that bavdegalutamide does not cause direct inhibition of CYP3A4 (34). The present study confirmed the in vitro findings, with no major drug–drug interactions observed between abiraterone and bavdegalutamide and no clinically significant effects on pharmacokinetics, suggesting that coadministration with bavdegalutamide should not require abiraterone dosing adjustments. Therefore, bavdegalutamide 420 mg once daily with abiraterone once daily in a 28-day cycle was chosen as the R2PD. In addition, preliminary clinical activity was observed in a subset of patients, particularly those with tumors harboring AR LBD mutations.

The median rPFS of 16.3 months from the start of bavdegalutamide treatment in the overall population seems relatively prolonged among patients who were already experiencing PSA progression on abiraterone. In the subgroup of patients who had mCRPC at abiraterone start, the median rPFS was 11.1 months, whereas in the patient subgroup with mHSPC at abiraterone start with PSA progression, the median rPFS was 16.3 months. In comparison, in the phase III placebo-controlled COU-AA-302 trial of first-line abiraterone treatment in chemotherapy-naïve patients with mCRPC, the rPFS from the start of abiraterone treatment was 16.5 months compared with 8.3 months in the placebo group (35). However, the reported benefit of second-line ARPI on radiographic progression after progression on first-line abiraterone treatment is more variable and can, for instance, be affected by a change in concomitant corticosteroid (36–39). Of note, no patients switched corticosteroids during this trial.

The common AR T878 and/or H875 LBD mutations have been shown to confer shorter real-world survival in patients with mCRPC after first-line novel hormonal agent ARPI initiation compared with matched patients whose tumors did not harbor these mutations [16.4 (95% CI, 6.9–not reached) versus 59.9 (95% CI, 43.5–not reached) months; P = 0.0007; ref. 19]. Consistent with results from a previous phase I study in which single-agent bavdegalutamide demonstrated promising antitumor activity in patients who had previously received AR antagonists, including enhanced activity observed in a subgroup of patients with tumors harboring AR T878 and/or H875 missense LBD mutations (24), we observed enriched clinical activity with bavdegalutamide in combination with abiraterone in patients harboring AR T878/H875 mutations, which are associated with resistance to AR-targeted therapy (40).

Numerous PROTACs are under clinical evaluation in solid tumors and hematologic malignancies (21, 30). The clinical trial data reported here advance our understanding of the potential utility of PROTACs in prostate cancer treatment, providing the first clinical evidence of safety and clinical activity of combining an AR PROTAC with an established therapy in abiraterone. Preclinical data suggest that PROTAC AR degraders have multiple advantages over small-molecule ARPIs for the treatment of mCRPC (21, 22). As previously reported, bavdegalutamide can target and degrade wild-type AR and clinically relevant AR mutants, outperforming enzalutamide in preclinical models, retaining activity in high-androgen settings, and blocking tumor growth in ARPI-resistant prostate cancer models (20, 22). By targeting multiple mechanisms of ARPI resistance, AR-targeted PROTACs and combinations not only have the potential to provide benefits after ARPI treatment failure but also early in the disease course, in which AR-targeted PROTAC/ARPI combinations may overcome emerging AR pathway resistance mechanisms. The findings reported here further support AR degradation with PROTACs as a therapeutic strategy for combination therapy in advanced prostate cancer.

Limitations of this study include the small study population and lack of a comparator arm, both characteristic of early-phase clinical trial design, necessitating further assessment of the clinical utility of this combination regimen in larger, late-phase clinical trials. Additionally, the heterogeneous patient population with respect to clinical disease state at abiraterone initiation prior to study entry, along with the study entry criteria of increasing PSA levels without known radiographic evidence of disease, limits the generalizability of the results and the interpretation of the clinical activity of bavdegalutamide. Because biopsies were not required, baseline AR signaling activity was not assessed, which may be informative for understanding treatment response in future trials. Finally, prior anticancer therapies and the racial and ethnic composition of the sample population could not be fully characterized due to missing data and reporting limitations from certain regions, restricting the interpretation of results across these demographic characteristics.

In conclusion, the combination of bavdegalutamide and abiraterone demonstrated a manageable safety profile and evidence of clinical activity in patients with mCRPC. These data further support the continued development of PROTAC AR degraders as monotherapy and in combination with standard-of-care ARPIs for the treatment of advanced prostate cancer via dual-targeted AR pathway inhibition.

Supplementary Material

Supplementary Data 1

Supplementary Appendices, Tables, and Figures.

Data Availability

The data generated in this study are not publicly available due to information that could compromise patient privacy or consent. Requests for data should be directed to the corresponding author. Data were generated by the authors and may be available upon reasonable request with the approval of the sponsor (Arvinas Androgen Receptor, Inc.; https://www.arvinas.com/). Any patient-level data will be anonymized, and study documents will be redacted to protect the privacy of trial participants.

Authors’ Disclosures

M. Smoragiewicz reports other support from Arvinas during the conduct of the study. A. Bernard-Tessier reports personal fees and nonfinancial support from Novartis and Bayer; personal fees from Merck, Janssen Pharmaceuticals, and Astellas Pharma; and grants from Servier outside the submitted work. M. Linch reports other from Bristol Myers Squibb, AstraZeneca, Janssen Oncology, BioNTech, Astellas Pharma, ADC Therapeutics, Bicycle Therapeutics, Pfizer/EMD Serono, and Merck Serono Research (consultant or advisory board); funding from Bristol Myers Squibb (institutional), AstraZeneca (institutional), and Shionogi (institutional); and travel, accommodations, and expenses support from Janssen Pharmaceuticals, Bristol Myers Squibb, MSD, and Merck. N.D. Shore reports ownership of Photocure stocks and consulting fees from Accord Healthcare, Alessa Healthcare, Amgen, Asieris Pharmaceuticals, Astellas Pharma, AstraZeneca, Aura Biosciences, Bayer, BioProtect, Bristol Myers Squibb, CG Oncology, Clarity Pharmaceuticals, Dendreon, Ferring Therapeutics, FIZE Medical, GlyTherix, ImmunityBio, Invitae, Janssen Pharmaceuticals, Lantheus, Eli Lilly and Company, MDxHealth, Merck, Minomic, Myriad, Novartis, Nusano, Pfizer, Photocure, PlatformQ Health, Promaxo, Protara Therapeutics, Sumitomo Pharma, Telix Pharmaceuticals, Tolmar, Tutelix, and UroGen Pharma. K.N. Chi reports grants from Arvinas during the conduct of the study, as well as grants and personal fees from Janssen Pharmaceuticals, AstraZeneca, Roche, Bristol Myers Squibb, Novartis, and Pfizer; personal fees from Astellas Pharma, Bayer, and Merck; and grants from Amgen outside the submitted work. A.O. Sokolova reports other support from Arvinas during the conduct of the study, as well as other support from AstraZeneca, Astellas Pharma, Bayer, and Pfizer outside the submitted work. S. Culine reports personal fees from Bayer, Johnson & Johnson, and Novartis outside the submitted work. M. McKean reports grants from Arvinas during the conduct of the study, as well as grants from Arcus Biosciences, Arvinas, Aulos Bioscience, Bayer, BioNTech, Boehringer Ingelheim, C4 Therapeutics, Dragonfly Therapeutics, Erasca, GenVivo, Gilead Sciences, Ikena Oncology, Immatics, Innate Pharma, Iovance Biotherapeutics, Jazz Pharmaceuticals, Kezar Life Sciences, Krystal Biotech, Nested Therapeutics, Novartis, NucMito Pharmaceuticals, OncoC4, Prelude Therapeutics, Replimune, Sapience Therapeutics, Synthrox, and TScan Therapeutics; grants and other support from Bicycle Therapeutics, Daiichi Sankyo, IDEAYA Biosciences, Immunocore Holdings, Merck, Moderna, Pfizer, Pierre Fabre, Regeneron Pharmaceuticals, and Teva Pharmaceuticals; and other support from AbbVie, Bristol Myers Squibb, Delcath Systems, Immunome, and Roche/Genentech outside the submitted work. M. Gleave reports other support from Astellas Pharma, AstraZeneca, AbCellera, Bayer, Janssen Pharmaceuticals, and Pfizer during the conduct of the study. M.E. Devitt reports personal fees from Onviv outside the submitted work. E. Zhi reports other support from Arvinas Inc. during the conduct of the study, as well as other support from Arvinas Inc. outside the submitted work. S. Landrette reports other support from Orphai Therapeutics outside the submitted work as well as a patent 11266654 issued to OrphAI Therapeutics, a patent 11166944 issued to Orphai Therapeutics, a patent 20210077500 pending to Orphai Therapeutics, a patent 20210030759 pending to Orphai Therapeutics, a patent 10799508 issued to Orphai Therapeutics, a patent 20200297706 pending to Orphai Therapeutics, a patent 10765682 issued to Orphai Therapeutics, a patent 20200268764 pending to Orphai Therapeutics, and a patent 20200253979 pending to Orphai Therapeutics. V. Patel reports other support from Arvinas Inc. during the conduct of the study, as well as other support from Arvinas Inc. outside the submitted work. X. Gao reports personal fees from 858 Therapeutics, Abeona Therapeutics, ADC Therapeutics, Arvinas, Bayer, Dendreon, Flare Therapeutics, Hinova Pharmaceuticals, Janssen Pharmaceuticals, KANAPH Therapeutics, Eli Lilly and Company, and Novartis outside the submitted work. No disclosures were reported by the other authors.

Authors’ Contributions

M. Smoragiewicz: Resources, writing–review and editing. A. Bernard-Tessier: Resources, data curation, supervision, investigation, writing–review and editing. M. Linch: Data curation, investigation, writing–review and editing. J. Shen: Conceptualization, resources, data curation, formal analysis, supervision, investigation, methodology, project administration, writing–review and editing. N.D. Shore: Conceptualization, data curation, investigation, writing–original draft, writing–review and editing. O. Parikh: Data curation, investigation, writing–review and editing. K.N. Chi: Resources, investigation, writing–review and editing. A.O. Sokolova: Data curation, project administration, writing–review and editing. S. Culine: Investigation, writing–review and editing. M. McKean: Data curation, investigation, writing–review and editing. M. Gleave: Resources, writing–review and editing. M.E. Devitt: Resources, investigation, project administration, writing–review and editing. E. Zhi: Formal analysis, validation, methodology, writing–review and editing. J. Gong: Investigation, writing–review and editing. S. Landrette: Conceptualization, methodology, writing–review and editing. M. Colasanto: Conceptualization, resources, data curation, formal analysis, supervision, validation, investigation, visualization, methodology, writing–original draft, project administration, writing–review and editing. V. Patel: Conceptualization, resources, data curation, formal analysis, supervision, validation, investigation, visualization, methodology, writing–original draft, project administration, writing–review and editing. X. Gao: Conceptualization, resources, data curation, formal analysis, supervision, investigation, methodology, project administration, writing–review and editing.