RESUMEN
Molecular diagnostics has drastically improved the survival rate of patients diagnosed with non-small cell lung cancer (NSCLC) over the last 10 years. Despite advancements in molecular testing, targeted therapies, and national guideline recommendations, more than half of NSCLC patients in the United States either never receive testing or patient care is not informed via molecular testing. Here, we sought to explore the relationship between DNA/RNA input, the molecular testing method, and test success rates. On a shared set of low-input reference test materials (n = 3), we ran both a hybrid capture-based, next-generation sequencing (NGS) assay and a multiplexed digital PCR (dPCR) panel. The dPCR panel was highly sensitive and specific for low-input samples in dilution studies ranging from 40 to 1 ng DNA and from 20 to 2.5 ng RNA, while NGS had up to an 86% loss in sensitivity as contrived sample inputs were serially diluted. The dPCR panel also demonstrated a high PPA (>95%) at diluted inputs as low as 15/7.5 ng DNA/RNA on 23 banked clinical samples with the same NGS hybrid capture assay at a high input. These data suggest that digital PCR is an accurate and effective way of identifying clinically relevant NSCLC mutations at low nucleotide input and quality.
RESUMEN
FDA approval of targeted therapies for lung cancer has significantly improved patient survival rates. Despite these improvements, barriers to timely access to biomarker information, such as nucleic acid input, still exist. Here, we report the analytical performance and concordance with next-generation sequencing (NGS) of a highly multiplexed research-use-only (RUO) panel using digital PCR (dPCR). The panel's analytical sensitivity and reactivity were determined using contrived DNA and RNA mixes. The limit of blank was established by testing FFPE curls classified as negative by pathology. Concordance was established on 77 FFPE samples previously characterized using the Oncomine Precision Assay®, and any discordant results were resolved with Archer Fusionplex® and Variantplex® panels. The analytical sensitivity, reported as the estimated mutant allele fraction (MAF), for DNA targets ranged from 0.1 to 0.9%. For RNA targets (ALK, RET, ROS, NTRK 1/2/3 Fusions, and MET Exon 14 skipping alteration), the analytical sensitivity ranged from 23 to 101 detected counts with 5 ng of total RNA input. The population prevalence-based coverage ranged from 89.2% to 100.0% across targets and exceeded 99.0% in aggregate. The assay demonstrated >97% concordance with respect to the comparator method.