Comprehensive tumor-agnostic evaluation of genomic and epigenomic-based approaches for the identification of circulating tumor DNA in early-stage breast cancer
Mitchell J. Elliott, J. Kim, A. Dou, Jesús Fuentes‐Antrás, Eitan Amir, Michelle B. Nadler, Emily Van de Laar, Celeste Yu, Raouf El Cheikh, Antonio Silvestro, Lillian L. Siu, Philippe L. Bédard, Hal K. Berman, David W. Cescon
Abstract
BACKGROUND: The detection of circulating tumor DNA (ctDNA) after curative-intent therapy, referred to as molecular/minimal residual disease (MRD), is prognostic of disease recurrence in early-stage breast cancer (EBC). Tumor-agnostic approaches that rely on mutation-based assessment in fixed panels of common cancer driver genes have shown limited utility for detecting MRD in EBC. Methylation-based MRD (mMRD) may overcome the limitations of genomic-based MRD (gMRD), though limited clinical validation is available. MATERIALS AND METHODS: To investigate this, we analyzed 290 longitudinally banked plasma samples from 95 participants diagnosed with early-stage estrogen receptor (ER)-positive/human epidermal growth factor receptor 2-negative (ER-positive) and triple-negative breast cancer (TNBC) undergoing neoadjuvant chemotherapy using a high-sensitivity genomic and epigenomic-based, tumor-agnostic ctDNA platform. RESULTS: The baseline (pre-chemotherapy) ctDNA detection (mMRD) rate was 72.5% (66/91) across all participants (ER-positive: 33/48, 69%; TNBC: 33/43, 77%). Baseline ctDNA detection (mMRD) was associated with a higher risk of recurrence [hazard ratio (HR) 9.4, 95% confidence interval (CI) 1.3-70.3, P = 0.03]. Detection of ctDNA (mMRD) in the post-operative and follow-up periods were prognostic of worse event-free survival (EFS) (HR 17.0, 95% CI 6.0-48.0, P < 0.0001) with 62.5% sensitivity and 100% specificity for recurrence (positive predictive value 100%). The median lead time from mMRD detection to clinical recurrence was 152 days (range 15-748 days). gMRD, derived from plasma-only panel-based next-generation sequencing, was evaluated in all matched time points; the prognostic value was limited by clonal hematopoiesis of indeterminate potential, including pathogenic mutations in common cancer driver genes. Despite refinements in gMRD analysis, it remained inferior to mMRD. A combination of mMRD and gMRD did not outperform mMRD alone. CONCLUSION: These results support further development of tumor-agnostic mMRD assays for the detection of ctDNA and assessment of these assays to develop clinical utility in this setting.