Abstract 3384: Evaluation of a tumor informed MRD assay with contrived breast cancer samples

Abstract

Abstract The non-invasive detection of circulating tumor DNA (ctDNA) from plasma has been shown to have clinical value for detection of minimal residual disease (MRD), emergence of resistance, and predicting treatment response. The sensitive detection of MRD following curative treatment allows for the identification of patients destined to recur. Higher assay sensitivities enable the longest lead times to clinical recurrence. Evaluations using several contrived pan-cancer samples have shown that the Inivata RaDaR tumor-informed personalized assay has a high sensitivity to detect ctDNA. The tumor-informed analysis utilizes sequencing of the tumor tissue to identify up to 48 clonal somatic variants, which are selected to generate a personalized panel to track the variants in patient plasma. To evaluate this technology in a disease specific manner, we generated a set of contrived plasma samples from commercial breast cancer patients and healthy individuals. The biological materials allowed us to generate a plasma sample with defined range of tumor content between 0.2 and 0.001% variant allele fraction (VAF). The sample dilutions were intended to recapitulate challenging cell-free DNA inputs that can be encountered in a clinical setting (median input: 5.76 ng [range 0.22 to 103.73]. Through our pan-cancer and breast cancer evaluations, we were able generate both sensitivity metrics along with a small limit of detection (LOD) study. The assay successfully detected tumor fractions in samples tested at 1X and 1.5X of the established LOD95% of 0.0011% VAF. RaDaR was able to detect the majority of MRD+ samples at 0.5X LOD. We will present data showing the level of sensitivity based on using both 2000 and 20000 genomic equivalents. The RaDaR assay showed high sensitivity for pan-cancer as well as ER+ breast cancer, which is known to have a low TMB compared to other subtypes. In this albeit small LoD study the Inivata assay was highly sensitive for the detection of ctDNA which is imperative to confidently detect ctDNA in advance of overt clinical recurrence. Citation Format: Hugh Russell, Paul LaBrousse, Daniel Stetson, Greg Jones, Giovanni Marsico, Tim Forshew, Robert McEwen, Elza De Bruin, Christopher Abbosh, Darren Hodgson, Brian Dougherty, James Hadfield, Carl Barrett. Evaluation of a tumor informed MRD assay with contrived breast cancer samples [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 3384.