Abstract 3505: Classification of cancer subtypes by cfDNA fragmentomics analysis

Abstract Introduction Fragmentation patterns of cell-free DNA (cfDNA) from whole genome sequencing offer insights into nucleosome occupancy, providing a tool to infer epigenomic and transcriptomic information. The nucleosome complex, the primary protector of cfDNA, is reflected in the size distribution of cfDNA fragments, with a mode fragment size of 167 bp corresponding to the wrapping of DNA around a single nucleosome. Nucleosome occupancy patterns at transcription factor binding sites (TFBS) exhibit differences between circulating tumor DNA in cancer patients and normal plasma controls, as well as among different tumor subtypes. Therefore, cfDNA fragmentomics can be employed for cancer subtyping analysis. Methods We utilized the Griffin framework to classify tumor subtypes based on nucleosome profiling of cancer-specific TFBS and tumor subtype-specific chromatin accessibility regions from low-pass whole genome sequencing data of cfDNA. For prostate cancer subtyping, AR and ASCL1 binding sites were used to distinguish between androgen receptor-dependent prostate cancer (ARPC) and neuroendocrine prostate cancer (NEPC). ER and ERBB2 were employed for breast cancer subtyping to differentiate between ER-positive and ER-negative tumor subtypes. Results Nucleosome profiling patterns at AR binding sites (ARBS) were compared among over 500 prostate cancer plasma samples and 42 normal plasma backgrounds. The ARBS nucleosome profiling abnormality score (ARBS score) was quantified by comparing its GC-corrected fragment central coverage with the normal background using a standard Z-score. Majority of prostate cancer samples with a high ARBS score could be classified as ARPC. For 12 samples with high tumor fraction but low ARBS score and no AR copy number amplification, we found that 8 had a high ASCL1 nucleosome profiling abnormality score. All these samples also exhibited RB1 copy number loss, suggesting that these 8 samples likely belong to NEPC. Nucleosome profiling of 171 breast cancer plasma samples at breast cancer-specific ER binding sites revealed that majority of samples had a high ERBS abnormality score, indicating likely ER-positive breast cancer. Conclusion Our results demonstrate that nucleosome profiling of tumor cfDNA exhibits distinct patterns across different tumor types. Integrating nucleosome profiling at cancer-specific transcription factor binding sites can facilitate cancer subtyping, highlighting its potential clinical utility for patient stratification. Citation Format: Lisha Zhu, Chao Dai, Shuang Gan, Shidong Jia, Pan Du. Classification of cancer subtypes by cfDNA fragmentomics analysis abstract. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts) ; 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84 (6Suppl): Abstract nr 3505.