Circulating free DNA derived from active chromatin as a predictive biomarker for clinical benefit to checkpoint inhibitor-based therapies in metastatic leiomyosarcoma.

Abstract

11539 Background: Leiomyosarcoma (LMS) is a common subtype of soft tissue sarcoma with a poor prognosis in the metastatic setting. LMS shows minimal benefit from monotherapy immune checkpoint inhibitors (CPI), however combinatorial CPI strategies may be effective in part due to tumor enrichment of epigenetic alterations. The DAPPER trial (NCT03851614) was a randomized, single center phase II study of durvalumab combined with olaparib or cediranib. Of the 30 LMS patients enrolled, 36.3% (n = 11) experienced disease stabilization or shrinkage. The present study aims to leverage a novel active chromatin cell-free DNA (cfDNA ac ) platform to investigate the epigenetic and genomic profiles of LMS patients in the DAPPER trial, with the goal of identifying biomarkers associated with clinical benefit from CPI-based therapies. Methods: Baseline plasma samples (n = 30) from LMS patients in the DAPPER trial were processed using a proprietary cfDNA ac capture assay that enriches active chromatin cfDNA. Following whole genome sequencing, univariate analysis and machine learning-based recursive feature selection were used to identify genomic features associated with clinical benefit rate (CBR, defined as RECIST v1.1 complete or partial response, or stable disease lasting > 6 months). Results: We identified 918 promoter and exon features that were significantly different (p < 0.01) at baseline and could segregate patients who achieved CBR from those who did not. Over-representation analysis of these gene features using Gene Ontology (p adj < 0.05) showed enrichment in biological pathways associated with double-strand break repair, inflammatory response, and immune response - specifically T-cell receptor activation and signaling, and macrophage homeostasis in patients with CBR. Conclusions: This study highlights the utility of cfDNA ac profiling as a non-invasive method for identifying biomarkers that predict clinical benefit from CPI-based therapy in patients with advanced LMS. Further analyses are ongoing to evaluate whether the genomic-derived features correlate with other clinical outcomes, such as progression-free survival, overall survival, and orthogonal data (e.g. tumor tissue RNA-seq).