Abstract 711: Co-detection of circulating tumor DNA and RNA for enhanced detection of minimal residual disease in patients with chemorefractory triple-negative breast cancer

Abstract

Abstract Background: Two-thirds of patients with early-stage Triple-Negative Breast Cancer (TNBC) will harbor residual disease (RD) in the breast after neoadjuvant chemotherapy (NAC) and are at a high risk-of-recurrence. We and others have established the detection of minimal residual disease (MRD) as flagged by circulating tumor DNA (ctDNA) to be associated with poor prognosis, though sensitivity remains a significant challenge. ctRNA has been seldom examined as a reservoir for somatic mutation detection, though it is abundant in the circulation. Herein, we hypothesize that co-detection of ctDNA/ctRNA will (1) improve the sensitivity of mutation detection in metastatic TNBC, and (2) robustly stratify risk for patients who have no evidence of disease. Methods: 6 and 36 patients with metastatic and early-stage TNBC were evaluable, respectively. For experimental co-detection, plasma nucleic acids were isolated, and custom cDNA synthesis and library preparation were performed using the Roche Avenio ctDNA platform in parallel with the standard pipeline followed by high-depth sequencing. Early-stage participants were sampled as part of a post-neoadjuvant clinical trial. Plasma samples were collected at the first day of adjuvant intervention. For detection of MRD, analytical sensitivity was defined as the percent of known recurrences in which ctDNA/ctRNA was detected, and specificity was defined in parallel with non-recurrences. Time-to-event analyses were conducted using the Log-Rank and Cox Regression methods. Results: In metastatic patients, co-detection across 12 mutations universally increased the number of mutated molecules detected (average increase: 64%; p=0.007) compared to ctDNA alone. 15/36 (42%) of early-stage patients had recurred at the 2-year primary trial endpoint. Index somatic mutations were detected in all RD tissues. ctDNA/ctRNA was co-detected in 10/15 recurrences (sensitivity = 67%). The longest lead time between mutation detection and clinical recurrence was 20.93 months (range=0.07 to 20.93). ctDNA/ctRNA was detected in 0/21 non-recurrences (specificity = 100%). The detection of ctDNA/ctRNA robustly stratified risk-of-recurrence (p<0.0001; median DFS (mos): 5.85 vs. NR; HR=13.04 (2.94 -57.90)). Discussion: Our customized method of ctDNA/ctRNA analysis detected significantly more mutated molecules in patients with overt disease and achieved robust sensitivity, specificity and risk stratification for early-stage patients. In particular, 67% sensitivity represents a significant improvement over a ctDNA-based study we performed in the same context, in which we observed only 31% sensitivity. The unplanned nature of this study is the most significant limitation from several perspectives, likely restricting the performance of the co-detection assay. Prospective performance validation versus ctDNA alone in this setting is essential downstream. Citation Format: Yu-Hsiang Chen, Bradley A. Hancock, Jeffrey P. Solzak, Bryan P. Schneider, Kathy D. Miller, Milan Radovich. Co-detection of circulating tumor DNA and RNA for enhanced detection of minimal residual disease in patients with chemorefractory triple-negative breast cancer [abstract]. In: Proceedings of the Annual Meeting of the American Association for Cancer Research 2020; 2020 Apr 27-28 and Jun 22-24. Philadelphia (PA): AACR; Cancer Res 2020;80(16 Suppl):Abstract nr 711.