Prospective ctDNA genotyping for treatment selection in metastatic castration-resistant prostate cancer (mCRPC): The Canadian Cancer Trials Group phase II PC-BETS umbrella study.
Edmond Michael Kwan,Moira Katherine Rushton,Wilson Tu,Sebastien J. Hotte,Som D. Mukherjee,Michael Ong,Michael Paul Kolinsky,Zineb Hamilou,Eric Winquist,Cristiano Ferrario,Robyn Jane Macfarlane,Fred Saad,Muhammad Salim,Di Maria Jiang,Dongsheng Tu,James Hutchenreuther,Matti Annala,Lesley Seymour,Kim N. Chi,Alexander William Wyatt
DOI: https://doi.org/10.1200/jco.2023.41.6_suppl.218
IF: 45.3
2023-02-20
Journal of Clinical Oncology
Abstract:218 Background: Precision oncology trials in mCRPC rely on genomic profiling of tumor tissue but testing failure rates are 30-40%. Incorporating liquid biopsy screening into trial designs may address limitations of tissue-only genotyping. We report findings from the first 500 plasma samples screened on Prostate Cancer Biomarker Enrichment and Treatment Selection (PC-BETS), a phase II multicenter, eight-arm Canadian umbrella trial (NCT03385655) using circulating tumor DNA (ctDNA) to match mCRPC patients to biomarker (BM)-informed targeted therapies. Methods: mCRPC patients previously treated with novel androgen receptor inhibitor therapy were eligible after PSA and/or radiological progression. Plasma cell-free DNA and matched leukocyte DNA underwent deep targeted sequencing with an exon-limited panel (Feb 2017-Sep 2020), or an expanded panel integrating select introns and a genome-wide copy number grid (July 2020-present). A molecular tumor board (MTB) assigned patients to treatment arms based on prespecified BM criteria (BM+), or by randomization if BM negative (BM-). The primary endpoint was clinical benefit rate (PSA50 response; RECIST CR/PR; or SD ≥12 weeks). We report tumor content (ctDNA%), genomic alterations, and BM status for the whole cohort. Results: As of Nov 2021, 503 samples were screened from 444 patients, with 496 passing quality control. 345 samples (70%) had ctDNA ≥1% (ctDNA+), of which the median ctDNA fraction was 20% (IQR 6-44%). 72% of ctDNA+ samples were BM+ (52% of all screened samples). Driver alterations influencing BM status included AR (76%; 59% gain, 24% mutation), PI3K pathway (37%; PTEN 30%, PIK3CA 6%, AKT 3%), and DNA repair defects (26%: mismatch repair 5%, BRCA2 7%, ATM 6%, CDK12 7%, other 6%). The expanded panel detected additional intronic structural variants in baseline ctDNA+ samples ( PTEN 1% vs 25%; BRCA2 0.6% vs 5%; AR 16% vs 37%), and identified whole genome doubling and segmental deletion events. To date, 167 patients have been enrolled to a substudy (83 BM+, 84 BM-). Median time from blood draw to MTB decision improved over time (first vs second half of screening period: 28 vs 17d) with implementation of optimized lab workflows, standardized genomic reports, and hierarchical genomic eligibility assessment. As of Sep 2022, 485 patients have been screened (updated results will be presented). Conclusions: Prospective centralized screening of ctDNA is feasible for guiding precision oncology initiatives. Improvements to assay design, robust availability of targeted therapies and an adaptive approach to biomarker assessment allowed high detection of actionable tumor alterations. Our framework can be used in future trials to stratify patients according to genomic alteration status. Study accrual to PC-BETS is ongoing, with a screening target of 600 patients. Clinical trial information: NCT02905318, NCT03385655 .
oncology