Comprehensive Validation of RNA Sequencing for Clinical NGS Fusion Genes and RNA Expression Reporting

PUBLICATIONS

March 10, 2021

AACR Annual Meeting 2021, Tempus-authored —

Tumor genome sequencing has emerged as a powerful tool for identifying biomarkers for targeted cancer therapies. While DNA sequencing is a well-established method and considered a gold standard, RNA sequencing (RNA-seq) can identify anomalies in gene transcription, regulation of gene expression, and gene fusions, which have critical diagnostic and therapeutic impacts. Tempus Labs is CLIA-certified, CAP-accredited and offers several clinically validated NGS assays for solid tumor and hematological malignancy testing, including the xT targeted 648-gene oncology panel and the xE whole-exome ~20,000 gene panel. The Tempus whole-transcriptome assay by hybrid-capture NGS has been enhanced to improve coverage, focusing on fusion genes and RNA expression calls. We clinically and analytically validated fusion calls for gene rearrangement detection and clinical reporting, and analytically validated RNA expression counts for research purposes only.

For RNA fusion (translocation) calling, we sequenced 96 tumor samples, including formalin-fixed, paraffin-embedded (FFPE) tissue, whole blood, and bone marrow samples from 12 cancer types; 82 of these samples contained reportable fusions. Of these, 86 samples were positive for RNA fusions, including 62 positive samples with targeted fusion breakpoints (i.e., utilizing custom probes designed to the breakpoints and spiked-in at hybridization) and 24 positive samples with untargeted fusion breakpoints (i.e., utilizing no custom probes). Targeted fusion accuracy was evaluated against the first version of the RNA sequencing assay. The first version of this assay was evaluated with cell line controls. In addition, validation included a specificity test and a rolling validation of fusions via DNA sequencing (positive controls).

The concordance for targeted fusions was 100%, with 99.9% sensitivity and 99.9% specificity. For untargeted fusions, the overall concordance was 97%, with 97% sensitivity. For RNA expression calling, we conducted a clinical linearity study using 88 samples and testing 18 genes to measure concordance between qPCR ΔCT values and normalized gene expression levels. Of these, 15/18 genes met the established acceptance criteria of R > 0.75.

In conclusion, the whole-transcriptome by hybrid-capture NGS assay offers clinically and analytically validated unbiased detection of common and novel gene rearrangements, as well as analytically validated gene expression data for comprehensive research analyses.

View the full publication here.

Authors: Jun Hu, Jerod Parsons, Brittany Mineo, Josh SK Bell, Jenna Malinauskas, Joshua Drews, Jack Michuda, Calvin McCarter, Rasika Dhond, Jack C. Tyndall, Nithya Sreenivasan, Sheeri Hanjra, Shannon Gallagher, Ankit Jambusaria, Naihui Zhou, Catherine Igartua, Nike Beaubier, Richard A. Blidner, and Robert Tell