Clinical performance of Signatera Genome assay in a cohort of patients (pts) with solid tumors.

Abstract

3142 Background: Circulating tumor DNA (ctDNA) has emerged as a powerful, minimally invasive biomarker of treatment response and pt prognosis. Signatera, a tumor-informed, mPCR-NGS ctDNA assay, offers high sensitivity and specificity for detecting molecular residual disease (MRD). Signatera (exome) uses its proprietary approach to select a highly curated set of tumor variants, followed by deep sequencing of plasma libraries at >100,000x per variant. Signatera Genome uses the same proven technology and may provide an advantage over exome in certain cases. However, beyond analytical improvements, it remains unclear if Signatera Genome provides superior performance and utility compared to the clinically validated exome-based version in the clinical setting. In this study, we assessed the clinical performance of the Signatera Genome assay in a cohort of pts with solid tumors. Methods: We performed a retrospective analysis of clinically annotated residual pt samples from commercial ctDNA testing (Signatera, exome-based, 16-plex mPCR-NGS assay). Adjuvant treatment decisions and ctDNA-cadence of testing were at the provider’s discretion. Signatera Genome assays were designed, consisting of 64 high-quality variants, from the respective pts’ matched tumor and normal whole genome sequencing data. These assays were used to detect ctDNA in the associated pts’ plasma utilizing a sample calling strategy that combined the target confidences and sample-level noise into a final confidence score. ctDNA concentration was measured in mean tumor molecules per mL of plasma (MTM/mL). Longitudinal plasma samples represented postoperative time points until recurrence/end of follow-up. The correlation between any time postsurgical ctDNA positivity and recurrence-free survival (RFS) was assessed using Cox regression analysis. Results: The Signatera Genome assay achieved a high analytical sample-level specificity of 99.8% (healthy subjects). Clinical performance was assessed in a real-world cohort of >300 pts with several cancer types, including breast cancer, non-small cell lung cancer (NSCLC), melanoma, and renal cell carcinoma (RCC). Among pts with relapse, the Signatera Genome assay detected ctDNA ahead of clinical recurrence as confirmed by imaging. Pts with postsurgical ctDNA-positivity demonstrated significantly inferior RFS compared to ctDNA-negative pts. This trend was consistent across all cancer types investigated. Multivariate analysis adjusted for tumor type and stage revealed ctDNA-positivity to be the most significant prognostic factor associated with RFS. Performance metrics by cancer type will be presented. Conclusions: Here we report the largest Signatera Genome ctDNA study to date across multiple solid tumor histologies. The data indicate robust performance and concordance with Signatera Exome. Prospective clinical trials are underway evaluating clinical utility.