Method development for pancreatic and ovarian cancer baseline ctDNA detection and measurable residual disease monitoring.

Background: Pancreatic cancer and ovarian cancer are very challenging to diagnose at early stages. The endoscopic retrieval of biopsy tissue from a suspected benign or malignant lesion is challenging due to the tissue's nature. Therefore, within the Instand-NGS4P framework, we developed Measurable Residual Disease (MRD) prototypes to analyze blood plasma samples, with the aim of cost-effectively supporting the differential diagnosis of suspected pancreatic or ovarian neoplasms. Methods: Our MRD prototypes examine blood plasma for mutations in cell-free DNA in specific genes associated with pancreatic neoplasms or ovarian neoplasms, respectively. Unique molecular identifiers (UMIs) are used to enable bioinformatic error correction. Ultra-deep sequencing is demonstrated on sequencing platforms from two different vendors (Illumina and MGI). We provide detailed information on bioinformatic processing of sequencing data to perform error-correction. Results: Using commercially available reference standards, we demonstrate stable mutation detection down to a variant allele frequency (VAF) of 0.1% with both 40 ng and 20 ng DNA input. At a coverage of 4,000x duplex consensus reads, only two false positives were observed, which can be efficiently mitigated using an appropriate filtering strategy. Conclusions: The technical usability of our MRD prototype has been clearly demonstrated for stable low-level VAF detection in commercial reference samples.