Abstract 692: Analytical validation of a tissue-free epigenomic assay for circulating tumor DNA (ctDNA)-based molecular residual disease (MRD) detection in early-stage cancer

Abstract Introduction: Molecular residual disease (MRD) detection can offer valuable prognostic insights and guide personalized treatment decisions. High sensitivity and specificity are required for accurate MRD assessment, with most methods currently relying on tissue sequencing for informing MRD assay design. However, tissue testing can be logistically challenging and not all patients have accessible tissue. Here, we describe the analytical validation of Guardant Reveal for plasma-only, epigenomic-based detection and quantification of ctDNA in CRC, lung, and breast cancers. Methods: Guardant Reveal evaluates signals in thousands of differentially methylated regions (DMRs) and classifies samples as ctDNA detected or not detected. Quantitative tumor fraction (TF) is estimated by normalizing selected cancer DMRs with matched control regions. Analytical sensitivity was assessed in titrations of cancer clinical samples and cell lines into non-cancer samples. Analytical specificity was assessed in samples from age-matched self-declared healthy donors. Qualitative precision was assessed in within- and between-batch replicates of ctDNA-positive and ctDNA-negative samples. Accuracy was assessed for concordance to a genomic and methylation-based reference assay. Results: Analytical sensitivity studies determined the ≥ 95% limit of detection (LoD) for each cancer type to be down to an order of 10-5-10-4 at both minimal (3-4ng) and typical (15ng) cfDNA input amounts. The analytical specificity (limit of blank) study using 102 age-matched healthy donor samples demonstrated 100% specificity (ctDNA negative results) for each cancer type caller. Within-batch and between-batch qualitative precision was 100% (81/81) for ctDNA-positive samples across cancer types, and 100% (36/36) for ctDNA-negative samples. In studies using titration replicates, TF demonstrated high quantitative precision and accuracy down to an order of 10-4, a range which encompasses 97% of ctDNA-positive clinical MRD samples. Comparison of ctDNA detection to a reference method demonstrated ≥97% positive percent agreement (PPA), and ≥95% negative percent agreement (NPA) for 153 CRC, 92 lung cancer, and 96 breast cancer samples. All discordant samples were borderline to the detection threshold. TF demonstrated quantitative equivalence to the reference method across the range of TF. Conclusions: Guardant Reveal demonstrated high analytical sensitivity, specificity, and TF quantitative precision in CRC, lung and breast cancer. This technology demonstrates the potential for ultra-sensitive and accurate MRD detection using a tissue-free assay, enabling precise monitoring for cancer recurrence. Citation Format: Katie Quinn, Amy Wilfert, Ranjani Lakshmin, Sai Chen, Albert Lee, Jun Zhao, Daniel Gaile, Yunpei Chang, Ruth McCole, Jordan Burke, Derek Dustin, Thereasa Rich, Jack Tung, Kristin Price, Darya Chudova. Analytical validation of a tissue-free epigenomic assay for circulating tumor DNA (ctDNA)-based molecular residual disease (MRD) detection in early-stage cancer [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 1 (Regular Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_1):Abstract nr 692.