We've just had news that Natera can transition Medicare patients to a genome-based MRD test. And we've got Caris doing an IPO as it emphasizes its genome/transcriptome flagship test.
See a new 21-page paper by Yudina, Lennerz, and colleagues, from Boston Gene. They report on clinical and analytical validation of a combined RNA DNA exome assay in a large tumor cohort - over 2000 tumors.
Find it here:
https://www.nature.com/articles/s43856-025-00934-3
See Lennerz' Linked-In summary:
##
AI Corner
Here's a Chat GPT summary and 10 take-aways.
##
Summary:
This study by Yudina, Lennerz, and colleagues presents the clinical and analytical validation of a combined RNA and DNA exome assay for cancer diagnostics using 2,230 tumor samples.
The integrated approach demonstrated significantly improved detection of actionable mutations, gene fusions, and expression profiles compared to DNA-only sequencing, with robust performance across multiple sample types including FFPE.
Key Conclusions (Emphasizing Clinical Impact and Surprises):
-
Combined DNA/RNA exome testing identified clinically actionable alterations in 98% of tumors, dramatically improving over traditional DNA-only assays.
-
RNA-seq rescued key mutations missed by DNA-seq alone, including low-allele-fraction mutations and gene fusions, altering potential treatment decisions.
-
The integrated assay uncovered complex genomic rearrangements, including cryptic gene fusions and allele-specific expression patterns that would be invisible with DNA-only panels.
-
RNA-based expression profiling and tumor microenvironment (TME) classification showed high reproducibility, supporting integration of immune signature diagnostics into routine pathology.
-
Fusion transcript detection sensitivity reached 100% in benchmark testing, outperforming popular tools like STAR-Fusion and Arriba, especially in low-expression contexts.
-
Even degraded FFPE samples yielded reliable results, owing to optimized capture-based protocols for both RNA and DNA—a major surprise in RNA assay feasibility from FFPE.
-
An exome-wide somatic variant reference set was developed and made publicly available, filling a major gap in clinical bioinformatics benchmarking.
-
Coverage thresholds were optimized: 150X for WES and ≥50M reads for RNA-seq, balancing cost and performance—important for labs considering implementation.
-
Gene signature reproducibility was excellent (CV < 3%), enabling consistent tumor subtype classification and immunotherapy prediction, even with input as low as 10 ng RNA.
-
The assay demonstrated stability and repeatability across multiple runs and reagent batches, making it suitable for longitudinal tumor profiling and standardized clinical deployment.
