Abstract 6628: Deep dive into the performance of structural variants reported by PGDx elio plasma focus Dx

Abstract Introduction: Liquid biopsy NGS assays allow for detection of genes clinically relevant to oncology including structural variants such as amplifications and translocations. Due to their low concentration in plasma, structural variants can be difficult to detect with liquid biopsy. Translocations in particular present challenges due to their low prevalence in ctDNA as opposed to RNA. PGDx elio™ plasma focus Dx is an IVD tumor profiling assay which reports 33 genes including difficult to detect structural variants with high sensitivity, specificity, accuracy, and reproducibility. Here, the performance of structural variants is explored using data from analytical validation studies. Methods: Samples were processed with PGDx elio plasma focus Dx, a NGS based tumor profiling assay capturing 33 clinically relevant genes using cfDNA obtained from plasma. For amplifications, 5 genes are reported (ERBB2, MET, CCND1, CD274, and FGFR2) and for translocations, 3 genes are reported (ALK, NTRK1, and RET). Libraries were sequenced with 150 bp paired end NGS and processed through the EPF software which performs demultiplexing, alignment, and variant calling utilizing DK for amplifications and PARE for translocations. Analytical validation studies were conducted at the gene level across key studies including accuracy, sensitivity, specificity, and reproducibility. For accuracy, results were compared to 5 orthogonal methods with a total of 931 clinical samples across 35 solid tumor types. Specificity was assessed with 68 healthy plasma samples. To assess sensitivity, all amplifications and translocations were established with a dilution series using cell line blends spiked into plasma and confirmed with clinical plasma blends diluted into normal plasma. Performance between clinical plasma and contrived samples was established with a contrived sample functional characterization study. Reproducibility was assessed with an interlaboratory reproducibility (ILR) study across 3 different sites with over 500 total replicates. Results: The test demonstrated variant concordance of 93.1% PPA and 99.8% NPA for translocations and 90.1% PPA and 100% NPA for amplifications (excluding borderline calls, 1.2-1.8-fold). Analytical specificity was 100% for both amplifications and translocations. Analytical sensitivity (limit of detection, (LoD) 95%) was assessed for each structural variant on the panel and ranged from 0.41% - 0.49% FRF for translocations and 1.3 - 1.7-fold for amplifications. The ILR study resulted in 96.1% average percent agreement (APA) and 99.6% average negative agreement (ANA) for amplifications and 95.3% APA and 99.5% ANA for translocations. Conclusions: These analytical validation results demonstrate PGDx elio plasma focus Dx detects structural variants with high accuracy, sensitivity, specificity, and reproducibility, supporting the use for precision oncology. Citation Format: Amy Meltzer, Jesse Fox, Kenneth C. Valkenburg, Tonya Watkins, Sonomi Dillon, Pengfei Gu, Ana Perez-Lebron, Robert Summersgill, Annabelle Culi, Nesha Jackson, Stephen Higgins, Kristy Waszkiewicz, Nobel Girmay, James Occean, Cal Palumbo, David Holcomb, Kelly Gerding, David Riley, Sam Angiuoli, Aanavi Karandikar, Rami Zahr, Jennifer Dickey, Eric Severson, Shakti Ramkissoon, Brian Caveney. Deep dive into the performance of structural variants reported by PGDx elio plasma focus Dx [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 6628.