Prevalence of androgen receptor ligand-binding domain mutations (AR-LBDm) in circulating tumor DNA (ctDNA) versus tissue biopsies in participants (pts) with metastatic castration-resistant prostate cancer (mCRPC) and other tumor types.
Abstract
3049 Background: AR-LBDm is a common mechanism of resistance to AR-directed therapies in patients with mCRPC. However, data on the utility of ctDNA-based versus tissue-based AR-LBDm detection in mCRPC and the prevalence of AR-LBDm in nonprostate cancers are lacking. We evaluated AR-LBDm in ctDNA and tumor tissue samples from pts with advanced solid tumors enrolled in various Merck & Co., Inc. trials. Methods: Samples came from pts with mCRPC and 25 other tumor types (eg, bladder, colorectal, ovarian, and pancreatic cancers). AR-LBDm data were available from ctDNA via 2 fixed panel-based ctDNA NGS assays or from tumor tissue samples (mainly archival) via a fixed-panel NGS assay. Data obtained from the 3 assays were nonoverlapping. AR-LBDm prevalence was evaluated in each tumor type and by BRCAm or HRRm status in pts with mCRPC. Results: The analysis included 3026 samples assessed by ctDNA assay 1 (mCRPC, n = 1785; other, n = 1241), 2232 samples assessed by ctDNA assay 2 (mCRPC, n = 378; other, n = 1854), and 8181 samples from tissue biopsies (mCRPC, n = 833; other, n = 7348). AR-LBDm was detected in 21.0% of pts with mCRPC by ctDNA assay 1 and 19.8% of pts with mCRPC by ctDNA assay 2. Across all other tumor types, only 1 pt (with hepatocellular carcinoma) had an AR-LBDm in ctDNA. The prevalence of selected AR-LBDm in tissue biopsies was 5.4% (45/833) in pts with mCRPC, 0.6% (1/159) in pts with salivary cancer, and 0% in all other tumor types. AR-LBDm prevalence in pts with mCRPC increased with later-line treatments (Tx) both by tissue and ctDNA analyses and was similar regardless of BRCAm or HRRm status (table). The prevalence of the AR-LBD T878A mutation (by ctDNA) in pts with mCRPC was higher after prior Tx with abiraterone than with enzalutamide (20.9% [43/206] vs 1.7% [3/173]). AR-LBDm was observed in tissues biopsied from the mCRPC setting (prevalence, 12.8% [12/94]) and was associated with higher AR transcriptional activity than in AR-LBD–negative tissues. Conclusions: Similar prevalence of AR-LBDm was observed in ctDNA of pts with mCRPC by 2 different ctDNA assays; AR-LBDm prevalence in archival tissue samples was lower than by ctDNA analysis, likely due to the Tx-emergent nature of AR-LBDm. AR-LBDm prevalence was similar regardless of BRCAm or HRRm status in mCRPC, although prevalence of certain AR-LBDm is impacted by the specific prior AR-directed Tx. These data support ctDNA-based (rather than tissue-based) AR-LBDm testing in pts with mCRPC. AR-LBDm prevalence, % (n/N) Frontline Tx Later-line Tx ctDNA assay 1 16.6 (122/734) 23.8 (252/1059) ctDNA assay 2 Not available 19.8 (75/378) Tissue biopsy 0.6 (3/535) 4.1 (147/3625) AR-LBDm +ve* AR-LBDm -ve * BRCA Mut WT 26.5 (9/34)27.5 (95/345) 73.5 (25/34)72.5 (250/345) HRR Mut WT 28.8 (30/104)26.9 (74/275) 71.2 (74/104)73.1 (201/275) *Per ctDNA assay 1.
Article Details
Journal Info
Journal of Clinical Oncology
Lippincott Williams & Wilkins
Authors (11)
Emmanuel S. Antonarakis
Masonic Cancer Center, University of Minnesota
Alice Bernard-Tessier
Department of Medical Oncology, Gustave Roussy, Villejuif, France
Cai Chen
Amir Vajdi
Merck & Co., Inc., Rahway, NJ
E.J. Dettman
Merck & Co., Inc., Rahway, NJ
Razvan Cristescu
13Merck & Co., Inc., Rahway, United States
Nicole Barkley
Merck & Co., Inc., Rahway, NJ
Carol Elaine Pena
Merck & Co., Inc., Rahway, NJ
April Wang
Merck & Co., Inc., Rahway, NJ
Evan Y. Yu
Fred Hutchinson Cancer Center, University of Washington, Seattle, WA
Karim Fizazi
Centre Oscar Lambret, University of Paris-Saclay, Lille, France