Sasha Main (1,2), Dr. Mitchell Elliott (1,3,4), Zhen Zhao (1), Althaf Singhawansa (1), Dr. Jinfeng Zou (1), Dr. Scott Bratman(1,2,5), Dr. David Cescon (1,3,4)
Affiliations:
1) Princess Margaret Cancer Centre, University Health Network, Toronto M5G 2C1, Ontario, Canada.
2) Department of Medical Biophysics, University of Toronto, Toronto M5G 1L7, Ontario, Canada.
3) Division of Medical Oncology and Hematology, Department of Medicine, University of Toronto, Toronto M5S 1A8, Ontario, Canada.
4) Institute of Medical Science, University of Toronto, Toronto M5S 1A8, Ontario, Canada.
5) Department of Radiation Oncology, University of Toronto, Toronto M5T 1P5, Ontario, Canada.
Background: CDK4/6 inhibitors (CDK4/6i), in combination with endocrine therapy, are the standard first-line treatment for patients with ER+/HER2- metastatic breast cancer. Acquired resistance to therapy is inevitable for patients, highlighting the need for novel biomarkers to predict non-responders and identify rational next lines of therapy. Analyzing circulating tumor DNA (ctDNA) through liquid biopsy offers a minimally invasive method for discovering biomarkers of CDK4/6i resistance and capturing disease heterogeneity at multiple metastatic sites. While most studies have primarily focused on genetic alterations in ctDNA, expanding the scope to include epigenetic features—such as methylation, hydroxymethylation, and nucleosome positioning—may uncover novel biomarkers of CDK4/6i resistance. Here, we assess the feasibility of simultaneously evaluating multiple epigenetic features with a single sequencing platform on ctDNA derived from preclinical models for CDK4/6i resistance biomarker discovery.
Methods: We applied methods to simulate ctDNA in tissue culture models, utilizing a nuclease treatment to mimic nucleosomal distributions found in plasma. Using matched CDK4/6 inhibitor (CDK4/6i) sensitive and resistant breast cancer cell lines (n=10), we generated simulated ctDNA samples. We conducted six-letter sequencing (6L-seq) on simulated ctDNA derived from a single initial cell line (CAMA-1), with biological replicates and shallow sequencing coverage (average 1.5X). We then evaluated paired RNA-seq, ATAC-seq, and simulated ctDNA WGS relative to simulated ctDNA 6L-seq.
