PS4-05-27: Longitudinal multi-omic profiling of tissues and blood to track the molecular evolution of metastatic breast cancer

Abstract Background Despite treatment advances, long-term survival for patients with metastatic breast cancer (mBC) remains low. While many tumors exhibit an initial response to conventional or targeted therapies, this is typically followed by the acquisition of resistance. While a subset of these acquired resistance events are driven by known mechanisms (e.g. ESR1 mutation on hormone therapy), others are driven by mechanisms that remain unsolved. Moreover, the trajectory and timeline by which tumors acquire resistance remains challenging to predict. To address this, we developed a study following the molecular evolution of mBC in tissues and circulating tumor DNA (ctDNA) combining both DNA sequencing and DNA methylation analysis. Methods We developed a protocol enrolling patients with a new diagnosis of de novo or first recurrence HER2-negative mBC. For enrolled patients, primary tumor and one or more metastatic sites are profiled using comprehensive genomic profiling (CGP) (ProvSeq Solid Tumor) and DNA methylation (DNAm) (Twist Human Methylome Panel). Patient blood samples are collected using Streck tubes at enrollment and at 12-week intervals for 3 years. Blood ctDNA is extracted and profiled using CGP (ProvSeq Liquid) and DNA methylation (Twist Human Methylome Panel). Mutations and methylation signatures are assessed in primary and metastatic tissue, and over time in ctDNA. Results While recruitment and longitudinal sample collection is ongoing, we report preliminary analyses from the first 40 participants enrolled in the study. Using CGP testing across solid tumor and liquid biopsy samples, we found 735 unique reportable variants, with 165 variants annotated as clinically-significant (OncoKB database), with 16 patients harboring a pathogenic alteration in PIK3CA. Solid tumor variant detection via CGP was highly accurate down to 5% variant allele fraction and down to 0.5% variant allele fraction in liquid CGP. For patients with both primary and metastatic tissue samples (n=20) mean concordance of variants was 86%. For variants deemed clinically significant in the tissue, 36% were also found in the closest ctDNA and 40% were found in ≥ 1 ctDNA sample over the monitoring period. Thus far, two patients (both receiving hormone therapy) have demonstrated ESR1 mutations. One was detected in the first ctDNA and progressed within 3 months. The other was detected 3 months after the diagnosis of metastatic disease, but did not progress clinically until 5 months later when the VAF had increased from 3% to 13%. Both solid tissue and liquid methylation analysis yielded high coverage across genome-wide CpG sites assessed in the panel, and current efforts are focused on investigating methylation changes in non-mutational acquired therapeutic resistance. Conclusions The combination of comprehensive genomic profiling and genome-wide methylation profiling is highly feasible in both tissue and liquid biopsies and may represent a synergistic strategy for maximizing diagnostic yield. Given the incomplete portrait of genomic evolution contributing to therapeutic response in mBC, our ongoing efforts seek to provide more comprehensive maps of mutational and epigenomic changes that may yield better treatment strategies in this population. Citation Format: H. Kaplan, A. Dowdell, J. Wagner, J. Welle, S. Reynolds, C. Carney, D. Page, A. Conlin, K. McCormick, F. Yan, E. Johnston, D. File, T. Wahl, J. Lopez, Z. Topp, K. Paulson, B. Xie, S. Montgomery, A. Bartlett, C. Bifulco, B. Piening. Longitudinal multi-omic profiling of tissues and blood to track the molecular evolution of metastatic breast cancer abstract. In: Proceedings of the San Antonio Breast Cancer Symposium 2025; 2025 Dec 9-12; San Antonio, TX. Philadelphia (PA): AACR; Clin Cancer Res 2026;32(4 Suppl):Abstract nr PS4-05-27.