Phase 1/2 study of ARV-110, an androgen receptor (AR) PROTAC degrader, in metastatic castration-resistant prostate cancer (mCRPC).

Abstract

17 Background: ARV-110 is a first-in-class, oral PROteolysis TArgeting Chimera (PROTAC) protein degrader that selectively targets AR. Patients (pts) with mCRPC have limited treatment (tx) options due to decreasing AR dependence of tumors upon successive therapies. Previous phase 1 data indicated clinical activity for ARV-110 in heavily pretreated pts with mCRPC and suggested enhanced activity in pts with specific molecular profiles, eg, AR T878 and H875 mutations, leading to a phase 2 expansion (ARDENT) to further characterize ARV-110 in biomarker-defined pt subgroups. We report results of the ongoing phase 1/2 study. Methods: In phase 1, pts with mCRPC and disease progression after ≥2 prior therapies (enzalutamide and/or abiraterone required) received ARV-110 orally once or twice daily (QD or BID) in sequential cohorts (3 + 3 dose escalation design). Primary objectives were to assess ARV-110 safety and select the recommended phase 2 dose (RP2D). Phase 2 pts with mCRPC and 1–2 prior novel hormonal agents (NHAs) ± chemotherapy were assigned to 3 biomarker-defined subgroups: 1) AR T878 and/or H875 mutations, 2) AR L702H mutation or AR-V7 (variants not degraded by ARV-110 in nonclinical studies), and 3) wild-type AR or other AR alterations. A fourth subgroup enrolled pts based on clinical history of less prior tx: ≤1 therapy for mCRPC, 1 NHA, and no chemotherapy. Primary objective is to assess ARV-110 antitumor activity. Results: As of Aug 26, 2021, 173 pts were enrolled (67 in phase 1; 106 in phase 2). In phase 1, ARV-110 doses ranged from 35–700 mg QD or 210–420 mg BID; 420 mg QD was selected as the RP2D based on safety, pharmacokinetics, and efficacy. Across 140 biomarker-evaluable phase 1/2 pts with ≥1 month of prostate-specific antigen (PSA) follow-up, 26 with AR T878A/S and/or H875Y mutations had best PSA declines ≥50% (PSA50) and ≥30% (PSA30) of 46% and 58%, respectively, vs 10% and 23% in 114 pts without these mutations. Of 7 RECIST-evaluable pts with AR T878A/S and/or H875Y mutations, 6 had tumor shrinkage (2 with confirmed partial responses), and 4 remain on tx. Five of 19 (26%) PSA-evaluable pts in the fourth subgroup (only 1 prior NHA; no prior chemotherapy) achieved PSA50. Overall PSA50 and PSA30 response rates were 16% and 29%, respectively. There were no grade ≥4 tx-related adverse events (TRAEs) in 113 pts treated at the RP2D. The most common any grade TRAEs at the RP2D were nausea (42%; grade 3: 1%), fatigue (27%; grade 3: 1%), vomiting (23%; grade 3: 1%), decreased appetite (19%; grade 3: 0), diarrhea (15%; grade 3: 2%), and alopecia (11%). Conclusions: ARV-110, a novel AR protein degrader, demonstrates clinical activity in a post-NHA, heavily pretreated mCRPC pt population, with greatest PSA50 activity and RECIST responses in pts with AR T878 and/or H875 mutations, likely representing a particularly ARV-110–sensitive population. ARV-110 merits further investigation in pts with mCRPC. Clinical trial information: NCT0388861.