Abstract PR004: Biological insights into tissue-agnostic plasma cfDNA methylation signature for surveillance of head and neck tumor recurrence
Bibliographic record
Abstract
Abstract BACKGROUND: We recently described a clinical validation of a tissue-agnostic genome-wide methylome enrichment molecular residual disease (MRD) assay for head and neck cancers (HNC) (Liu, ESMO 2024). Herein, we investigate the cfDNA methylation signature utilized by this test to demonstrate the biological relevance with respect to the tissue of origin and epigenetic mechanisms of dysregulation that alter normal cellular activity. We evaluate the signature to gain insights and to characterize the biology of this biomarker and to investigate the application of cfMeDIP-seq assay as a diagnostic tool. METHODS: From unblinded training set of 163 HNC patients, including HPV+ (n = 76), HPV- (n = 43), and HPV-unknown (n = 44) individuals, a cfDNA methylation classifier signature was developed to detect recurrence (n MRD+ = 62, n MRD- = 101) in samples from blood draws approximately at 3, 12, and 24 months post curative intent treatment using cfMeDIP-seq assay. Clinical performance of the signature was then successfully validated in a blinded HNC validation cohort (n = 162), including HPV+ and HPV- patients (Liu, ESMO 2024). Here, the signature’s ability to detect ctDNA was assessed using HNC tumor tissue (n = 523), adjacent normal tissue (n = 45), and healthy peripheral blood leukocytes (n = 93) DNA methylation data from TCGA. The signature’s ability to provide biological insights was assessed by evaluating enrichment in CpG elements, epigenetic regulators, regulatory pathways, and other biological pathways. RESULTS: Identified differentially methylated regions (DMRs) were significantly hypermethylated in HNC tumor tissue samples when compared to adjacent normal and healthy peripheral blood leukocytes (P = 1.3e-131), and the mean signature signal within tumor tissue samples correlated with tumor purity (P = 2.47e-21). These DMRs were also significantly enriched in CpG islands, shores, and shelves and could classify HNC patients into identifiable molecular subtypes. Further analysis of the DMRs, based on their genomic coordinates, shows association with known cancer genes and significant enrichment in pathways known to be associated with HNC, for example, epithelial to mesenchymal transition (P = 1.76e-4), and multiple functional pathways involved in DNA and transcription factor binding, transcriptional regulation, and chromatin remodeling. CONCLUSIONS: These results, along with the previous clinical validation, provide strong evidence that the cfDNA methylation signature developed in HNC detects ctDNA shed from residual tumor and provide insights into relevant tumor biology and epigenetic mechanisms, including CpG island hypermethylation and epithelial to mesenchymal transition. Furthermore, these data support the generalization of biomarker detection utilizing cfMeDIP-seq to a broad spectrum of indications. Citation Format: Yulia Newton, Justin Burgener, Margaret Gruca, Collin Melton, Jun Min, Abel Licon, Scott Bratman, Alan Williams, Daniel D De Carvalho. Biological insights into tissue-agnostic plasma cfDNA methylation signature for surveillance of head and neck tumor recurrence [abstract]. In: Proceedings of the AACR Special Conference: Liquid Biopsy: From Discovery to Clinical Implementation; 2024 Nov 13-16; San Diego, CA. Philadelphia (PA): AACR; Clin Cancer Res 2024;30(21_Suppl):Abstract nr PR004.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.001 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.001 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.002 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".