Whole genome sequencing-powered ctDNA sequencing for breast cancer detection
Isaac García-Murillas, Charles W. Abbott, Rosalind Cutts, S.M. Boyle, J. Pugh, Kathleen C. Keough, B. Li, R M Pyke, Fábio C. P. Navarro, Richard Chen, Kevin Dunne, Catey Bunce, S. Johnston, Alexander Ring, Simon Russell, AA Evans, A. Skene, I. E. Smith, N.C. Turner
Abstract
BACKGROUND: Circulating tumour DNA (ctDNA)-based detection of molecular residual disease (MRD) presents a strategy to identify patients at high risk of relapse. In this article, we profile early breast cancer patients with an ultrasensitive, whole genome sequencing (WGS)-based, tumour-informed ctDNA platform. MATERIALS AND METHODS: We analysed 617 plasma samples (median 8, range 2-14) from 78 patients (23 triple-negative breast cancer, 35 human epidermal growth factor receptor 2-positive, 18 hormone receptor-positive, and 2 unknown). Samples were collected at diagnosis before therapy, cycle 2 of neoadjuvant chemotherapy, post-surgery after neoad'juvant therapy if administered, every 3 months during the first year, and every 6 months thereafter. Plasma DNA was analysed using the NeXT Personal MRD platform, a tumour-informed WGS approach to produce personalized ctDNA sequencing panels tracking a median of 1451 variants per patient. MRD detection was correlated with clinical outcomes. RESULTS: ctDNA was detected at levels ranging from 2.19 parts per million (PPM) to 204 900 PPM (median 405 PPM), with 39% of all ctDNA detections in the ultra-low range <100 PPM. Of patients with samples at diagnosis, 98% (49/50) had ctDNA detected before treatment. At a median follow-up of 76 months (range 5-118 months), detection of ctDNA was associated with high risk of future relapse (P < 0.0001; log-rank test) and shortened overall survival (P < 0.0001) with a median lead time from ctDNA detection to clinical relapse of 15 months (range 0.9-61.5 months). MRD was identified in 100% (11/11) of patients who relapsed, with a median level of ctDNA at first MRD detection of 13.1 PPM. No ctDNA-undetected patients relapsed throughout follow-up (64/64). Comparison with exome-powered MRD detection assays showed improved sensitivity and lead time. CONCLUSIONS: A whole genome-powered MRD assay detected breast cancer relapse with a long lead time over clinical relapse, and was strongly associated with relapse-free survival. Rates of ctDNA detection at diagnosis were higher than those reported with exome-based tumour-informed assays.