Validation of a next-generation sequencing assay for the detection of EGFR mutations in cell-free circulating tumor DNA.

Detection of EGFR mutations from blood plasma represents a gentle, non-invasive alternative to rebiopsy and can therefore be used for therapy monitoring of non-small-cell lung cancer (NSCLC) patients. The aim of this project was to investigate whether the Reveal ctDNA™ 28 NGS assay (ArcherDX, Boulder, CO), has a comparable sensitivity and specificity to droplet digital PCR (ddPCR, gold-standard) and is therefore suitable for therapy monitoring of progressing lung cancer patients. First, we validated the NGS assay with a commercially available reference material (SeraCare, Massachusetts, US). Using an input of 22 ng, a sensitivity of 96% and a specificity of 100% could be achieved for variant allele frequencies (VAF) of 0.5%. For variants at a VAF of 0.1% the sensitivity was substantially reduced. Next, 28 plasma samples from 16 patients were analyzed and results were compared to existing ddPCR data. This comparative analysis of patient samples revealed a concordance of 91% between NGS and ddPCR. These results confirm that the Reveal ctDNA™ 28 NGS assay can be used for therapy monitoring of patients under TKI therapy. However, due to the slightly superior sensitivity of ddPCR, a combination of NGS (with broad coverage of a large number of genomic loci) and ddPCR (with targeted highly sensitive detection of specific mutations) might be the ideal approach.

Clinical Implementation of a Noninvasive, Multi-Analyte Droplet Digital PCR Test to Screen for Androgen Receptor Alterations.

Alterations of the androgen receptor (AR) are associated with resistance to AR-directed therapy in prostate cancer. Thus, it is crucial to develop robust detection methods for AR alterations as predictive biomarkers to enable applicability in clinical practice. We designed and validated five multiplex droplet digital PCR assays for reliable detection of 12 AR targets including AR amplification, AR splice variant 7, and 10 AR hotspot mutations, as well as AR and KLK3 gene expression from plasma-derived cell-free DNA and cell-free RNA. The assays demonstrated excellent analytical sensitivity and specificity ranging from 95% to 100% (95% CI, 75% to 100%). Intrarun and interrun variation analyses revealed a high level of repeatability and reproducibility. The developed assays were applied further in peripheral blood samples from 77 patients with advanced prostate cancer to assess their feasibility in a real-world scenario. Optimizing the reverse transcription of RNA increased the yield of plasma-derived cell-free RNA by 30-fold. Among 23 patients with castration-resistant prostate cancer, 6 patients (26.1%) had one or a combination of several AR alterations, whereas only 2 of 54 patients (3.7%) in the hormone-sensitive stage showed AR alterations. These findings were consistent with other studies and suggest that implementation of comprehensive AR status detection in clinical practice is feasible and can support the treatment decision-making process.