Cell-Specific DNA Methylation Markers in Plasma cfDNA Reveal Diagnostic Potential for Head and Neck Cancer

Head and neck cancer ranks as the seventh most common type of cancer worldwide. For patients diagnosed at advanced stages, the five-year survival rate remains below 40%, underscoring the urgent need for early detection strategies. However, current diagnostic approaches face considerable limitations. Furthermore, tissue biopsy remains invasive and impractical for routine screening. In light of these challenges, this study aims to identify and validate plasma cell-free DNA (cfDNA) methylation markers capable of enabling auxiliary assessment and facilitating prognosis assessment in head and neck cancer.

To achieve this goal, we first analyzed whole-genome bisulfite sequencing (WGBS) data from 205 samples covering 39 healthy tissue types. This comprehensive analysis was conducted to establish a baseline methylation landscape and identify methylation sites specifically associated with head and neck cells. Subsequently, differentially methylated CpG sites were screened by comparing head and neck tissues with other healthy tissues. To enhance the robustness and specificity of potential markers, more than 5 adjacent cell-specific methylated CpG sites located within 150 base pairs of each other were clustered into composite methylation markers, thereby reducing background noise and improving signal detection. These candidate markers then underwent a stringent specificity test to confirm their exclusive enrichment in head and neck tissues. Finally, one of the most promising markers was experimentally validated in plasma cfDNA samples collected from head and neck cancer patients compared with healthy controls, using targeted bisulfite sequencing to assess their detection performance in liquid biopsies.

Through comprehensive analysis of WGBS data from healthy tissue samples, a total of 1,157 tissue-specific methylation sites were identified, including 146 hypermethylated and 1,011 hypomethylated sites exclusive to head and neck tissues. By clustering adjacent CpG sites within a 150-base-pair window, these were consolidated into 11 composite methylation markers, each representing a distinct genomic region with coordinated methylation patterns. Subsequent specificity testing confirmed that all composite markers exhibited strong cell-type specificity, with minimal background methylation signals detected across other tissue types. This high specificity underscores their potential for accurate tissue-of-origin inference. Validation experiments in plasma cfDNA samples further demonstrated that methylation fragments corresponding to these composite markers were consistently detectable in cancer patients but entirely absent in healthy individuals.

This study systematically characterizes cell-specific methylation signatures of head and neck tissues and provides preliminary evidence for the feasibility of utilizing composite cfDNA methylation markers for non-invasive cancer detection. By integrating multiple adjacent CpG sites into composite markers, our approach achieves enhanced specificity and robustness compared to conventional single-site markers, providing a strategy to overcome limitations associated with tumor heterogeneity and low cfDNA abundance in early-stage disease. These findings not only provide a methodological framework for cfDNA-based auxiliary assessment but also offer promising candidate biomarkers for tracing the tissue of origin in head and neck cancer. This work is limited by a small sample size and pending further investigation in larger, prospective cohorts. However, this work illustrates the potential of cell-specific methylation analysis to contribute to improved disease monitoring and tissue-of-origin determination.