Transcriptionally Informed Nucleosome Profiling of Circulating Cell-Free DNA Predicts Breast Cancer Recurrence

Area of Science:

• Oncology
• Genomics
• Molecular Biology

Background:

• Cell-free DNA (cfDNA) analysis provides a minimally invasive method to monitor cancer progression.
• Understanding genomic and epigenetic changes in cfDNA is crucial for early cancer detection and recurrence prediction.

Purpose of the Study:

• To investigate the utility of cfDNA profiling for detecting breast cancer recurrence.
• To analyze cfDNA genomic and epigenetic alterations associated with therapy resistance and relapse.

Main Methods:

• Targeted sequencing of 26 gene loci in cfDNA from 150 breast cancer patients (105 primary, 45 recurrent).
• Analysis of cfDNA variant counts, fragment lengths, and fragmentation profiles.
• Application of machine learning to integrate cfDNA features for relapse prediction.

Main Results:

• Recurrent breast cancer samples showed increased cfDNA variant counts and shorter fragment lengths.
• Distinct cfDNA fragmentation patterns, including amplifications (e.g., ERBB2) and reductions (e.g., RERE, SYNPO2), were observed in recurrent samples.
• Nucleosome occupancy scores derived from RERE and SYNPO2 accurately distinguished recurrent from primary cancer (AUC = 0.826).
• Integrated cfDNA features achieved high accuracy in predicting breast cancer relapse via machine learning.

Conclusions:

• cfDNA-based profiling targeting transcriptional alterations is a sensitive strategy for breast cancer recurrence detection.
• Analysis of cfDNA fragmentation and associated epigenetic changes offers valuable insights into cancer relapse mechanisms.
• This approach holds potential for non-invasive monitoring of breast cancer patients.