Abstract BACKGROUND: mRCC is characterized by high intratumoral heterogeneity, posing challenges for the development of predictive or prognostic biomarkers. Plasma cell free DNA (cfDNA) may reflect the molecular landscape of metastatic disease and systemic anti-tumor response. Emerging data suggest that transcriptional signatures may subtype mRCC with distinct clinicopathologic phenotypes (Cancer Cell, PMID: 33157048). In this study, we estimate the differences in TFBS accessibility using fragmentation patterns in total cfDNA in healthy individuals and mRCC patients (pts). METHODS: Pts with mRCC undergoing immune checkpoint inhibitor (ICI) -based systemic treatment were recruited between June 2018 and December 2021 (NCT03702309). Blood was collected at pre-ICI and on-ICI (between 4-8 weeks) timepoints. Clinical outcomes were last updated on November 1, 2025. Blood samples from healthy controls were obtained from 18 volunteers. Plasma cfDNA whole genome sequencing (WGS) was performed with 10x coverage and aligned to GRCh38. The ratio of short (90-150 bp) over long (151-220 bp) DNA fragments were calculated in 100kb bins across the genome. GC content and read depth were corrected for short and long fragments. Genome-wide fragmentation profiles were generated using an adapted version of DELFI (Nature, PMID: 31142840). Fragment coverage analysis at TFBS was performed using the Griffin platform to infer chromatin accessibility (Nat. Commun. , PMID: 36463275). All statistical comparisons were adjusted for multiple testing with a false discovery rate set at 0. 05. RESULTS: A total of 27 pts had blood samples collected pre-ICI, of whom 24 also had matched samples on-ICI. 25 (93%) had clear cell histology. ICI regimens included: ipilimumab/nivolumab (n=15), nivolumab monotherapy (n=9), and pembrolizumab/axitinib (n=3). Compared to healthy individuals, pre-ICI samples of mRCC pts had a higher short fragment ratio (p0. 001). The short fragment ratio did not differ between pre-ICI and on-ICI samples. Fragment coverage analysis at 310 individual TFBS were performed for each cfDNA sample. Compared to healthy individuals, pre-ICI samples had 214 TFBS with significantly increased coverage (inferred to be nucleosome bound), and 9 TFBS with significantly decreased coverage (inferred to be open chromatin). There was no statistically significant difference in TFBS coverage in pre-ICI versus on-ICI samples in the entire cohort or within treatment subgroups. Among pts who received ipilimumab/nivolumab, higher coverage at the HOXB13 TFBS on pre-ICI cfDNA was associated with an increased hazard ratio (HR) for PFS (HR = 7. 7, p=0. 01) and a trend towards poorer OS (HR = 4. 6, p=0. 09). CONCLUSIONS: Fragmentomics-estimated TFBS accessibility was different between healthy individuals and pts with mRCC. Griffin analysis of cfDNA should be further assessed as a diagnostic tool for mRCC or as a measure of minimal residual disease in future studies. Validation of TFBS coverage as a predictive biomarker is warranted in larger cohorts. Citation Format: Changsu L. Park, Yi Yue Jiang, Lisa-Monique Edward, Jeffrey Bruce, Ming Han, Brooke Wilson, Aaron R. Hansen, Trevor J. Pugh, Lillian L. Siu, Pavlina Spiliopoulou. Fragmentomics-derived profiling of transcription factor binding site (TFBS) accessibility from liquid biopsies in metastatic renal cell carcinoma (mRCC) abstract. In: Proceedings of the AACR Special Conference in Cancer Research: Innovations in Kidney Cancer Research: From Molecular Insights to Therapeutic Breakthroughs; 2026 Mar 13-16; Philadelphia, PA. Philadelphia (PA): AACR; Cancer Res 2026;86 (5Suppl₂): Abstract nr A039.
Park et al. (Fri,) studied this question.