Detection of BRCA1, BRCA2, and ATM Alterations in Matched Tumor Tissue and Circulating Tumor DNA in Patients with Prostate Cancer Screened in PROfound

Abstract

Purpose: Not all patients with metastatic castration-resistant prostate cancer (mCRPC) have sufficient tumor tissue available for multigene molecular testing. Furthermore, samples may fail because of difficulties within the testing procedure. Optimization of screening techniques may reduce failure rates; however, a need remains for additional testing methods to detect cancers with alterations in homologous recombination repair genes. We evaluated the utility of plasma-derived circulating tumor DNA (ctDNA) in identifying deleterious BRCA1, BRCA2 (BRCA), and ATM alterations in screened patients with mCRPC from the phase III PROfound study. Experimental Design: Tumor tissue samples were sequenced prospectively at Foundation Medicine, Inc. (FMI) using an investigational next-generation sequencing (NGS) assay based on FoundationOne®Liquid to inform trial eligibility. Matched ctDNA samples were retrospectively sequenced at FMI, using an investigational assay based on FoundationOne®Liquid CDx. Results: 81% (503/619) of ctDNA samples yielded an NGS result, of which 491 had a tumor tissue result. BRCA and ATM status in tissue compared with ctDNA showed 81% positive percentage agreement and 92% negative percentage agreement, using tissue as reference. At variant-subtype level, using tissue as reference, concordance was high for nonsense (93%), splice (87%), and frameshift (86%) alterations but lower for large rearrangements (63%) and homozygous deletions (27%), with low ctDNA fraction being a limiting factor. Conclusions: We demonstrate that ctDNA can greatly complement tissue testing in identifying patients with mCRPC and BRCA or ATM alterations who are potentially suitable for receiving targeted PARP inhibitor treatments, particularly patients with no or insufficient tissue for genomic analyses.

Detection of BRCA1, and BRCA2 Alterations in Matched Tumor Tissue and Circulating Cell-Free DNA in Patients with Prostate Cancer in a Real-World Setting

“…These data are consistent with the prior studies [ 14 , 18 , 19 ]. Tukachinsky and colleagues reported that with tumor tissue as reference, the positive percentage agreement for BRCA1/2 on cfDNA was 93% [ 14 ].…”

Towards clinical implementation of circulating tumor DNA in metastatic prostate cancer: Opportunities for integration and pitfalls to interpretation

“…15-20% of mCRPC harbor deleterious HRR-related gene mutations, most commonly in BRCA2, ATM and CDK12. Concordance between tissue and ctDNA for HRR gene mutations is excellent, exceeding 90% in most studies (23,64,71,78,79). However, an HRR-related gene mutation alone may be insufficient to confer therapeutic vulnerability.…”

“…These resourcing requirements mean that a majority of clinicians around the world continue to have limited access to ctDNA profiling to inform patient care. Instead, clinicians typically encounter ctDNA testing in the clinical trial setting, where it is increasingly used alongside tumor tissue testing to screen for genomic alterations that may confer sensitivity to established and novel targeted therapies (20)(21)(22)(23)(24)(25). Outside of clinical trials, US FDAapproved commercial platforms are available (e.g., FoundationOne Liquid CDx, Guardant360 CDx) (26, 27), but can be cost prohibitive for patients unable to acquire partial or full financial reimbursement for testing.…”

“…However, plasma ctDNA screening is increasingly used to complement or replace existing screening procedures. Recent PARP inhibitor studies in mCRPC have endorsed ctDNA analysis alongside tumor tissue testing to cross-validate presence of HRR alterations (20)(21)(22)(23). PC-BETS (IND.234) and ProBio are two ongoing biomarker-directed prospective mCRPC clinical trials utilizing cfDNA testing for treatment allocation (25,93).…”

“…Many commercial and academic plasma cfDNA assays are capable of detecting somatic mutations below 1% ctDNA, but heterozygous and homozygous deletions typically require far greater ctDNA to be reliably detected ( 78 , 82 , 83 ). Therefore, low ctDNA% may be a limiting factor in detecting HRR alterations in cfDNA, especially large structural rearrangements and homozygous deletion events ( 23 ).…”