Epigenetic alterations, particularly DNA methylation, appear early in carcinogenesis, even earlier than physical lesions detectable by imaging. Because these signatures are stable and detectable in body fluids (blood, urine, stool), they can serve as excellent, non-invasive liquid biopsy markers for early diagnosis, disease monitoring, and screening. Liquid biopsy is a non-invasive and highly sensitive cancer screening approach that detects the methylation patterns of circulating tumor DNA or circulating cell-free DNA (cfDNA) in blood, urine, saliva, or feces. Technologies such as flow cytometry and next-generation sequencing (NGS), mass spectrometry-based approaches, droplet digital PCR (ddPCR), particularly circulating tumor cell analysis (enriched by flow cytometry), cfDNA sequencing, and emerging epigenetic and fragmentomic profiling are in different phases of development that have the potential to revolutionize the approach to detect and treat cancers^[2-4]. Likewise, advances in fragmentomics and histone modification profiling are revealing new dimensions of tumor biology that can be captured through liquid biopsy, allowing clinicians to monitor treatment response and measure minimal residual disease (MRD) with unprecedented precision.

Since epigenetic alterations are not permanent modifications to the DNA sequences, they can be druggable, making them a key focus for cancer therapy. The so-called epidrugs aim to reset the gene expression program of tumor cells and restore the activity of tumor suppressor genes or inhibit oncogenes by targeting epigenetic marks. Epidrugs include histone deacetylase inhibitors (HDACi), DNA methyltransferase inhibitors (DNMTi), histone methyltransferase inhibitors (HMTi), and bromodomain and extra-terminal motif protein inhibitors (BETi)^[7]. By inhibiting "writers" (such as DNMTs), "erasers" (such as HDACs), or "readers" (such as BET proteins), abnormally silenced tumor suppressor genes can be reactivated, or oncogenes can be inactivated.

Prognostics is the clinical assessment of a patient's treatment outcomes, including the probability of cancer progression, recurrence, or survival. Epigenetic markers can more accurately predict a patient's cancer progression, therapy response, survival and relapse risk than traditional staging^[7].

For example, O⁶-methylguanine-DNA methyltransferase (MGMT) promoter methylation acts as a critical biomarker in GBM. Specifically, silencing MGMT gene due to promoter methylation leads to decreased DNA repair capacity, enhanced sensitivity to temozolomide (TMZ) chemotherapy, and results in improved overall survival^[22]. Thus, MGMT promoter methylation serves as a key predictor of response to TMZ chemotherapy in GBM.

Clinically, Graded Prognostic Assessment (GPA) for brain metastases^[23] or Palliative Prognostic Score (PaP)^[24] are validated tools which assign numerical risk scores based on factors such as functional status (Karnofsky Performance score) and disease extent. The GPA is used to estimate survival and guide treatment for patients with brain metastases, while the PaP is used to predict 30-day survival in advanced cancer patients.

Multi-omics integration, computational models, and artificial intelligence (AI) contribute to the development of novel strategies for early cancer detection, risk stratification, MRD monitoring, and precise prognostics. Computational approaches are revolutionizing the field, with deep learning frameworks now capable of predicting metastatic potential from primary tumor methylation data, while AI-driven deconvolution techniques enable the dissection of tumor methylation profiles to uncover the immune and stromal components of the tumor microenvironment, which is key to predicting treatment response. cfDNA methylation signatures have emerged as highly sensitive biomarkers for MCED offering a promising route for non-invasive screening across multiple tumor types. Similarly, advances in fragmentomics and histone modification profiling are revealing new dimensions of tumor biology that can be performed through liquid biopsy, allowing clinicians to monitor treatment response and detect MRD with unprecedented precision. Cell-free chromatin state tracing (cf-EpiTracing) may provide early diagnosis, molecular subtyping and prognostic prediction by revealing disease origin and therapy responses^[25].

CRISPR-based epigenome editing will provide a powerful experimental toolkit to validate candidate regulatory regions and epigenetic drivers identified in large-scale datasets. Tang et al.^[26] systematically reviewed the advances using CRISPR-based molecular biosensing for pathogen detection and early tumor screening. They summarized the technological evolution of device integration, including automated and closed microfluidics to portable systems to achieve point-of-care testing.

Mapping the epigenetic regulatory networks that drive cancer initiation, progression, and therapy resistance remains a critical challenge. In gastrointestinal cancer, Huang et al.^[27] systematically illustrated the epigenetic dysregulation for providing hints in clinical translation. G-quadruplexes (G4s) are non-canonical secondary structures formed by guanine-rich DNA or RNA strands by Hoogsteen hydrogen bonds that help control gene replication, transcription, and RNA translation. DHX36, a member of the DEAD/H box superfamily, is a main helicase that unwinds G4s of target genes^[28,29]. Abnormal G4 regulation by helicases should be linked to cancer pathways and progression epigenetically, making them targets for new diagnostics, prognostics, and treatments^[30,31]. The DHX36-m6A-G4 axis might play roles in health and disease in an epigenetic way^[31,32].

Single-cell multi-omics (such as scRNA-seq + scATAC-seq) combined with bulk data will dissect the intratumoral heterogeneity and identify epigenetically distinct cell states. Li et al.^[33] reviewed precision management in chronic myeloid leukemia by single-cell sequencing. Furthermore, advanced techniques like ATAC-seq and ChIPmentation allow for rapid assessment of chromatin accessibility, providing high-precision, personalized diagnostics. Emerging technologies such as third-generation sequencing and spatial epigenomics now enable direct and spatial measurements of DNA methylation and chromatin accessibility, opening new possibilities for correlating epigenetic signatures with tissue architecture.

The field of pharmacoepigenomics is rapidly advancing, leveraging epigenetic markers to predict response to targeted epigenetic therapies and guide combination treatment strategies. Computational epigenetics can accelerate the identification of new therapeutic candidates by exploiting known epigenetic signatures and drug-epigenome interactions.