Analytical validation and clinical performance of a plasma-based CGP assay, NEO | PanTracer LBx.

Abstract

e15047 Background: Comprehensive genomic profiling (CGP) assays targeting hundreds of genes are increasingly used over smaller panels in patients with advanced solid tumors. CGP by liquid biopsy (LBx) complements traditional tissue-based CGP, mitigating tissue constraints, heterogeneity, and wait times. Here we describe the analytical validation (AV) and clinical performance characteristics of NEO | PanTracer LBx. As part of the clinical validation (CV), we report initial results across advanced stage cancers and compare concordance with an in-house amplicon-based 37-gene lung cancer panel (InVisionFirst-Lung) and other LBx CGP assays. Methods: Based on TSO500 chemistry, PanTracer LBx is a pan-solid tumor, next-generation sequencing assay designed to detect single nucleotide variants (SNVs) and small insertions/deletions (Indels), copy number variants (CNV), and fusions across, 517, 59 and 23 genes, respectively, as well as blood tumor mutational burden (bTMB) and microsatellite instability (MSI). Key assay performance characteristics (Limit of Detection, LoD; Limit of Blank, LoB; precision, analytical sensitivity/specificity) were determined for the above alteration types. Clinical validation is underway, testing specimens from patients with advanced cancers, from the same timepoint, with PanTracer LBx and InVisionFirst-Lung (n = 145 pts) or commercially available pan-tumor LBx CGP assays (n = 57 pts). Initial comparisons were limited to pathogenic variants in genomic regions with mutual coverage. Results: LoD90 assessment utilized SeraSeq complete ctDNA reference material at different input concentrations to detect known somatic variants present at different %VAF levels (0.1%-2%). LoD90 was estimated to be 0.17% for SNVs, 0.31% for Indels, 0.5% for fusions and 1.23x for CNVs. LoB for SNVs, Indels, CNVs and fusions, assessed by calculating the cumulative per gene specificity across 21 healthy donors tested, was 0%. To assess the analytical sensitivity and specificity, patient samples were sequenced and compared to results obtained with InVisionFirst -Lung. Positive percent agreement (analytical sensitivity) and negative percent agreement (analytical specificity) were 96.9% and 99.3%, respectively. Precision (inter-assay, -lot and -operator) was > 95%. CV confirmed high concordance rates between PanTracer LBx and LBx CGP assays with 92% agreement (47/51 variants). Conclusions: Performance evaluation of PanTracer LBx met all pre-specified AV acceptance criteria. Leveraging a large panel targeting key classes of somatic alterations across multiple genes, PanTracer LBx demonstrates results consistent with a smaller, highly sensitive targeted panel and similar large CGP panels. Additional CV is ongoing; further analysis will include additional samples, tumor-specific findings, cross-test comparisons, and particular biomarker and immune signature findings (MSI and bTMB).