Predicting immunotherapy benefit in leiomyosarcoma through active chromatin cfDNA profiling

Abstract Leiomyosarcoma (LMS) is an aggressive soft-tissue sarcoma characterized by epigenomic dysregulation and variable responsiveness to checkpoint inhibitors (CPIs). In this study, we evaluated a novel liquid biopsy platform based on circulating cell-free DNA active chromatin (cfDNA ac ) profiling to identify baseline biomarkers associated with clinical benefit rate (CBR) in 30 LMS patients treated with durvalumab plus olaparib or cediranib. A total of 1,570 molecular features across genomic regions were identified, and recursive feature elimination was applied to select discriminative cfDNA ac signatures. Patients achieving CBR demonstrated enrichment of pathways related to cell death, interferon-γ signaling, and immune activation. Signatures reflecting B-cell activation, T-cell activation, and extracellular matrix organization were significantly associated with improved progression-free survival (PFS; p < 0.05). In contrast, a baseline tumor fraction >5% was negatively associated with CBR (p = 0.041) and correlated with inferior PFS (p = 0.008). Distinct copy number variation profiles further characterized patients with CBR and were significantly associated with PFS. In this analysis, cfDNA ac profiling represents a promising non-invasive strategy to predict clinical benefit from CPI-based therapy in LMS. Prospective validation in independent cohorts is warranted to clarify its clinical utility.