Machine learning guided signal enrichment for ultrasensitive plasma tumor burden monitoring

In solid tumor oncology, circulating tumor DNA (ctDNA) is poised to transform care through accurate assessment of minimal residual disease (MRD) and therapeutic response monitoring. To overcome the sparsity of ctDNA fragments in low tumor fraction (TF) settings and increase MRD sensitivity, we previously leveraged genome-wide mutational integration through plasma whole genome sequencing (WGS). We now introduce MRD-EDGE, a composite machine learning-guided WGS ctDNA single nucleotide variant (SNV) and copy number variant (CNV) detection platform designed to increase signal enrichment. MRD-EDGE uses deep learning and a ctDNA-specific feature space to increase SNV signal to noise enrichment in WGS by 300X compared to our previous noise suppression platform MRDetect. MRD-EDGE also reduces the degree of aneuploidy needed for ultrasensitive CNV detection through WGS from 1Gb to 200Mb, thereby expanding its applicability to a wider range of solid tumors. We harness the improved performance to track changes in tumor burden in response to neoadjuvant immunotherapy in non-small cell lung cancer and demonstrate ctDNA shedding in precancerous colorectal adenomas. Finally, the radical signal to noise enrichment in MRD-EDGE enables de novo mutation calling in melanoma without matched tumor, yielding clinically informative TF monitoring for patients on immune checkpoint inhibition. ### Competing Interest Statement DAL, AJW, CCK, JB and MS submitted two patent applications. AS receives research funding from AstraZeneca, has served on Advisory Boards for AstraZeneca, Blueprint Medicines, and Jazz Pharmaceuticals, and has been a consultant for Genentech. MAP has received consulting fees from BMS, Merck, Array BioPharma, Novartis, Incyte, NewLink Genetics, Aduro, Eisai, and Pfizer, has received honoraria from BMS and Merck, and has received institutional support from RGenix, Infinity, BMS, Merck, Array BioPharma, Novartis, and AstraZeneca. CLA reports collaborations with C2i Genomics and Natera. MKC has received consulting fees from BMS, Merck, InCyte, Moderna, ImmunoCore, and AstraZeneca and receives institutional support from BMS. ST is funded by Cancer Research UK (grant reference number A29911); the Francis Crick Institute, which receives its core funding from Cancer Research UK (FC10988), the UK Medical Research Council (FC10988), and the Wellcome Trust (FC10988); the National Institute for Health Research (NIHR) Biomedical Research Centre at the Royal Marsden Hospital and Institute of Cancer Research (grant reference number A109), the Royal Marsden Cancer Charity, The Rosetrees Trust (grant reference number A2204), Ventana Medical Systems Inc (grant reference numbers 10467 and 10530), the National Institute of Health (U01 CA247439) and Melanoma Research Alliance (Award Ref no 686061). ST has received speaking fees from Roche, Astra Zeneca, Novartis and Ipsen. ST has the following patents filed: Indel mutations as a therapeutic target and predictive biomarker PCTGB2018/051892 and PCTGB2018/051893. JDW is a Consultant for Amgen; Apricity; Ascentage Pharma; Arsenal IO; Astellas; AstraZeneca; Bicara Therapeutics; Boehringer Ingelheim; Bristol Myers Squibb; Chugai; Daiichi Sankyo, Dragonfly; Georgiamune; Idera; Imvaq; Kyowa Hakko Kirin; Maverick Therapeutics; Psioxus; Recepta; Tizona; Trieza; Trishula; Sellas; Surface Oncology; Werewolf Therapeutics. JDW receives Grant/Research Support from Bristol Myers Squibb; Sephora. JDW has Equity in Tizona Pharmaceuticals; Imvaq; Beigene; Linneaus, Apricity, Arsenal IO; Georgiamune; Trieza; Maverick; Ascentage. DAL received research support from Illumina, Inc. DAL is a scientific co-founder of C2i Genomics.