Rapid, ultra-low coverage copy number profiling of cell-free DNA as a precision oncology screening strategy in metastatic castration-resistant prostate cancer (mCRPC)

144

Background: Although cell free DNA (cfDNA) profiling by next generation sequencing (NGS) holds great promise for precision oncology, the inability to estimate tumor content a priori typically results in ultra-deep sequencing (e.g. 10,000x) based approaches—often at limited loci—to ensure accurate assessment of cfDNA samples where tumor content can be < 0.1%. However, a large subset of patients with advanced cancers, where most precision oncology is applied, have much greater cfDNA tumor content. Methods: Here we demonstrate the utility of a pan-cancer, rapid, inexpensive, whole genome NGS of cfDNA approach (PRINCe) on benchtop sequencers as a precision medicine screening strategy based on ultra-low coverage (~0.005x) tumor content determination through genome-wide copy number alteration (CNA) profiling using 48 plasma cfDNA samples from patients with advanced cancer. Results: Using this approach, we identified therapeutically relevant focal CNAs in 13 of 48 (27%) cfDNA samples from patients with metastatic cancer, including 11 of 36 (31%) from patients with mCRPC. We further show that PRINCe is effective at whole genome coverages as low as 0.005x. Combining PRINCe with targeted multiplexed-PCR NGS of the same cfDNA enables mutation and CNA assessment using effective coverage as low as 25x based on calibrating sequencing depth to cfDNA tumor content. Lastly, PRINCe identifies pre-treatment cfDNA copy number profiles that can be used for inexpensive disease monitoring. Conclusions: Taken together, our screening approach enables broadly applicable cfDNA based precision oncology for patients with advanced cancer through rapid, inexpensive identification of therapeutically relevant CNAs and pre-treatment genomic profiles for disease monitoring, while also guiding additional cfDNA profiling approaches to reserve costly ultra-deep approaches for patients with very low cfDNA tumor content.