De novo characterization of cell-free DNA fragmentation hotspots boosts the power for early detection and localization of multi-cancer

Non-random cell-free DNA fragmentation is a promising signature for cancer diagnosis. However, its aberration at the fine-scale in early-stage cancers is poorly understood. Here, we developed an approach to de novo characterize the cell-free DNA fragmentation hotspots from whole-genome sequencing. In healthy, hotspots are enriched in gene-regulatory elements, including open chromatin regions, promoters, hematopoietic-specific enhancers, and, interestingly, 3end of transposons. Hotspots identified in early-stage hepatocellular carcinoma patients showed overall hypo-fragmentation patterns compared to healthy controls. These cancer-specific hypo-fragmented hotspots are associated with genes enriched in gene ontologies and KEGG pathways that are related to the initiations of hepatocellular carcinoma and cancer stem cells. Further, we identified the fragmentation hotspots at 297 cancer samples across 8 different cancer types (92% in stage I to III), 103 benign samples, and 247 healthy samples. The fine-scale fragmentation level at most variable hotspots showed cancer-specific fragmentation patterns across multiple cancer types and non-cancer controls. With the fine-scale fragmentation signals alone in a machine learning model, we achieved 48% to 95% sensitivity at 100% specificity in different early-stage cancer. We further validated the model at independent datasets we generated at a small number of early-stage cancers and healthy plasma samples with matched age, gender, and lifestyle. In cancer-positive cases, we further localized cancer to a small number of anatomic sites with a median of 80% accuracy. The results highlight the significance of de novo characterizing the cell-free DNA fragmentation hotspots for detecting early-stage cancers and dissection of gene-regulatory aberrations in cancers.