574 Minimally invasive NGS approach for the detection and follow-up of endometrial cancer patients

Introduction/Background Background: Minimally invasive biomarkers are a necessity in in a clinical oncology setting the clinical context. In this regard endometrial carcinomas (EC) are lacking. Although usually associated with a good prognosis, advanced EC patients show extremely poor clinical outcomes. Especially relevant in this clinical context is the lack of methods that allow for a proper and dynamic monitoring follow-up in EC. The aim of the present study is to evaluate the value of a minimally invasive NGS approach on longitudinal samples from advanced EC patients. Methodology Peripheral blood samples (n=32) from 18 patients with EC disease were collected at surgery (n=12), at progression disease (PD) (n=10) after first line systemic treatment and at sequential progressions (n=10). Total cfDNA was isolated from 5mL of plasma using QIAamp DNA Circulating Nucleic Acid Kit. Isolated cfDNA was quantified using Qubit and integrity using the TapeStation. After quality control the cfDNA was sequenced using the Oncomine™ Pan-Cancer cell free assay with an input ranging from 4.63 to 58.24ng (Median=20.23ng). Results Overall, 71.88% of samples (23/32) showed at least one pathogenic mutation with the variant allelic frequency ranging from 0.1% to 67.89% (Median=1.53%). Most patients presented mutations in TP53, PIK3CA and/or GNAS. Importantly, ctDNA dynamics reflect disease evolution with levels of ctDNA increasing with the development of PD. The study of longitudinal cfDNA samples with the NGS panel in EC allowed us to understand how the mutational landscape of patients behaves throughout time, with the arisen of novel mutations not found on the baseline samples, that may derive from treatment pressure. Conclusion The study of longitudinal cfDNA samples with the NGS cell-free panel represents a comprehensive strategy for a non-tumour informed characterization of endometrial tumours that reflects the clinical evolution and even allows for the identification of novel mutations that could be used as actionable targets. Disclosures The authors declare no conflict of interests.