Circulating tumor cell RNA biomarker for bavdegalutamide (ARV-110) in metastatic castration-resistant prostate cancer without AR T878/H875 mutations.

Abstract

235 Background: Bavdegalutamide (bavdeg; formerly ARV-110) is a first-in-class oral proteolysis targeting chimera (PROTAC) protein degrader that selectively targets androgen receptor (AR). A phase 1/2 study of bavdeg showed clinical activity in patients with metastatic castration-resistant prostate cancer (mCRPC) after 1-2 prior AR-targeted therapies, particularly in those harboring AR T878X/H875Y mutations. Although some patients with wildtype AR showed clinical benefit from bavdeg, no biomarker exists to select this subgroup of patients. Here we evaluated a circulating tumor cell (CTC) RNA biomarker for its potential to identify patients with mCRPC who may benefit from bavdeg in the phase 1/2 study despite the absence of AR T878X/H875Y mutations. Methods: Patients with mCRPC and disease progression after >1 prior novel hormonal agents were enrolled in the phase 1/2 study of bavdeg and consented for prospective collection of pre-treatment blood for CTC analysis at a single institution. CTCs were isolated using a negative selection microfluidic CTC enrichment device and analyzed for RNA expression of a panel of prostate cancer genes ( AGR2, FAT1, FOLH1, HOXB13, KLK2, KLK3, STEAP2, TMPRSS2, AR-V7 ) using a multiplex droplet digital PCR assay. A composite gene expression score (CTCm; Miyamoto et al. Cancer Discov. 2018) was compared to radiographic response at 2 months, PSA response, and duration of time on bavdeg. A cut-off value for CTCm was determined by receiver operating characteristic curve analysis. Association of CTCm with clinical outcomes was analyzed by two-sided Fisher’s exact test. Results: Twenty patients treated with bavdeg doses ranging from 280-700 mg daily and 140-420 mg twice daily consented to pre-treatment CTC RNA expression analysis. Circulating tumor DNA analyses revealed none of the patients harbored AR T878 or H875 mutations. Six patients remained on bavdeg >6 months with clinical benefit, while 14 discontinued therapy within 6 months. Radiographic responses at 2 months were stable, progressive, and not evaluable in 10, 7, and 3 patients, respectively. A best PSA decline ≥50% (PSA50) and ≥30% (PSA30) was achieved in 3 and 4 patients, respectively. Pre-treatment CTCm score <165 was associated with stable disease at 2 months (p = 0.015) and time on bavdeg >6 months (p = 0.014). CTCm did not correlate with PSA50 (p>0.9) or PSA30 (p=0.60). Presence of AR-V7 did not correlate with radiographic response (p=0.15), PSA50 (p>0.9), PSA30 (p=0.60), or time on therapy >6 months (p=0.14). Conclusions: Pre-treatment CTCm score <165 was associated with freedom from radiographic progression at 2 months and time on bavdeg >6 months in this small study mCRPC patients without AR T878/H875 mutations. Further investigation of the CTCm biomarker is warranted in mCRPC patients treated with bavdegalutamide and potentially other AR PROTACs.