ERASE-Seq Demonstrates Superior NGS Concordance to Digital PCR in ctDNA Liquid Biopsies of Lung Cancer Clinical Samples

Cristian Ionescu-Zanetti1, Nicholas Kamps-Hughes1, Jeff Jensen1, Jessica Garcia2, Lea Payen-Gay2 1Fluxion Biosciences, Alameda, CA 2Clinical Hospital Lyon, Lyon, France


A key performance hurdle in liquid biopsy implementation is detection of variants at extremely low allele frequencies (AF), typically to 0.1% AF and below. High sensitivity digital PCR methods such as ddPCR and BEAMing have demonstrated the ability to detect variants at these allele frequencies, but the techniques are limited to small panels. Broad panels are critical for diagnostic approaches such as tumor mutational burden, sub-clonal detection, and molecular pathway analysis.

Next-generation sequencing (NGS) removes the multiplexing limit and can test for thousands of variants in a single assay, but NGS lacks the necessary sensitivity. To address this we introduce ERASE-Seq (Elimination of Recurrent Artifacts and Stochastic Errors), a method for accurate and sensitive detection of ultralow frequency DNA variants. ERASE-Seq differs from previous methods by creating a variant event matrix (Fig. 1) that includes multiple sample runs and control data. Statistical analysis is applied at each position in the sequence, yielding a confidence score for every call. This approach identifies true positives with high sensitivity without introducing false positives.

ERASE-Seq was first tested on 8 samples comprising two different controls with variants at allele frequencies ranging from 0.1%-1%. ERASE-Seq showed 100% sensitivity with 0 false positives, testing across a 133-variant target list selected to include the most clinically significant variants. For clinical testing, plasma was isolated from blood samples of 11 lung cancer patients. Samples were tested by NGS and ddPCR or BEAMING. Duplicate NGS samples were evaluated using a custom caller and ERASE-Seq. NGS runs analyzed using the custom caller showed a combined 46% concordance in calling variants identified by ddPCR and BEAMing. ERASE-Seq demonstrated a concordance of 92% to ddPCR and BEAMing, and identified an additional 7 variants that were consistent with disease pathology.