Reducing cost and complexity in tumor-informed MRD by combining patient probe panels using Enspyre.

Abstract

e14533 Background: Tumor-informed minimal residual disease (MRD) assays track patient-specific variants, achieving higher sensitivity than tumor-naïve approaches, but requiring custom probe panels for each patient. For high-throughput laboratories, buying, storing, and handling tens or hundreds of thousands of custom panels brings significant cost and complexity, while little or no negative control data are generated for each panel, limiting assay specificity. Combining multiple probe panels reduces cost, as one large panel is cheaper than multiple small, and alleviates operational complexity. It also provides negative control data, since most probes in a given panel should be negative for each single patient. However, in standard MRD assays this approach dramatically increases the sequencing required due to capture of DNA from regions not of interest for a specific patient. Enspyre enriches variant-containing molecules prior to sequencing, meaning that probes targeting variants absent in a sample are inactive and do not result in significant unwanted DNA. This presents an opportunity to combine probe panels without increasing sequencing requirements, bringing significant benefits in performance, cost, and complexity. Methods: Five cell lines were used to assess the effect of combining panels, with a sixth used to mimic background cfDNA. Panels of 2,000 probes were designed to target variants present in each of the five cell lines but not the sixth, and were tested individually and in combination against a serial dilution of each cell line. Limit of detection was estimated for each panel using probit regression-based analysis, and compared to the limit of detection obtained using the combined pool on each cell line. The impact of pooling probe panels on sequencing depth requirements was assessed through downsampling of sequence reads in silico . Results: Consistent ‘ultra sensitive’ ctDNA detection at or below 10ppm was achieved across individual probe panels and using the combined 10,000 probe panel for all targeted cell lines. Sequencing depth requirements were not significantly increased through pooling, with ultra sensitivity achieved using as few as 10M read pairs (vs 100-500M read pairs using standard hybridization-capture sequencing). Conclusions: This work confirms that Enspyre’s variant enrichment enables combinatorial use of multiple probe panels without associated increases in sequencing depth requirements. In real-world settings this could have significant impact, enabling reductions in cost of custom probes and the logistical complexity of storing and tracking vast numbers of bespoke panels. Further, pooling of five 2,000 probe panels into a single pool results in up to 8,000 ‘known negative’ probes for each patient tested, providing substantial negative control data for each individual probe without incurring any additional burden, solving a significant challenge in the field.