Detection of Circulating Tumor DNA with a Single-Molecule Sequencing Analysis Validated for Targeted and Immunotherapy Selection

Abstract

Introduction: Comprehensive genetic cancer profiling using circulating tumor DNA has enabled the detection of National Comprehensive Cancer Network (NCCN) guideline-recommended somatic alterations from a single, non-invasive blood draw. However, reliably detecting somatic variants at low variant allele fractions (VAFs) remains a challenge for next-generation sequencing (NGS)-based tests. We have developed the single-molecule sequencing (SMSEQ) platform to address these challenges. Methods: The OncoLBx assay utilizes the SMSEQ platform to optimize cell-free DNA extraction and library preparation with variant type-specific calling algorithms to improve sensitivity and specificity. OncoLBx is a pan-cancer panel for solid tumors targeting 75 genes and five microsatellite sites analyzing five classes of NCCN-recommended somatic variants: single-nucleotide variants (SNVs), insertions and deletions (indels), copy number variants (CNVs), fusions and microsatellite instability (MSI). Circulating DNA was extracted from plasma, followed by library preparation using SMSEQ. Analytical validation was performed according to recently published American College of Medical Genetics and Genomics (ACMG)/Association for Molecular Pathology (AMP) guidelines and established the limit of detection (LOD), sensitivity, specificity, accuracy and reproducibility using 126 gold-standard reference samples, healthy donor samples verified by whole-exome sequencing by an external College of American Pathologists (CAP) reference lab and cell lines with known variants. Results were analyzed using a locus-specific modeling algorithm. Results: We have demonstrated that OncoLBx detects VAFs of ≥ 0.1% for SNVs and indels, ≥ 0.5% for fusions, ≥ 4.5 copies for CNVs and ≥ 2% for MSI, with all variant types having specificity ≥ 99.999%. Diagnostic performance of paired samples displays 80% sensitivity and > 99.999% clinical specificity. Clinical utility and performance were assessed in 416 solid tumor samples. Variants were detected in 79% of samples, for which 87.34% of positive samples had available targeted therapy.