Cancer Cell. 2025 Oct 02. pii: S1535-6108(25)00398-8. [Epub ahead of print]
The analysis of cell-free DNA (cfDNA) fragmentation patterns, known as "fragmentomics," has opened new opportunities in noninvasive cancer diagnostics. Due to its close relationships with genomic organization and cell death, cfDNA fragmentomics lies at the intersection of many aspects of cancer biology, including epigenetic dysregulation, transcriptomic alterations, and aberrant cellular turnover patterns. Recent advances in library preparation, sequencing technologies, and integrative epigenomic-fragmentomic analyses have uncovered novel fragmentomic features that reveal specific cellular dysfunctions in cancer. Additionally, cutting-edge artificial intelligence algorithms now harness high-dimensional fragmentomic features, boosting the precision and power of cancer detection. Promising results from recent clinical trials evaluating the utility of fragmentomic analyses in real-world settings support its potential. In this review, we explore the exciting frontiers of cfDNA fragmentomics, discuss critical unanswered questions, and highlight future directions to unlock the promise of fragmentomics-based liquid biopsies in cancer care.
Keywords: artificial intelligence; cancer detection; cancer risk prediction; cancer treatment monitoring; cell-free DNA; fragmentomics; machine learning