THE ROLE OF CFDNA, CTDNA, METHYLATION AND FRAGMENTOMICS IN EARLY CANCER DETECTION: A SYSTEMATIC REVIEW

Abstract

Background: Early cancer detection is a major strategy for reducing cancer-related morbidity and mortality. Circulating cell-free DNA (cfDNA) provides a minimally invasive substrate for detecting tumor-associated quantitative, genetic, epigenetic, and structural abnormalities. This systematic review evaluates the roles of total cfDNA, circulating tumor DNA (ctDNA), DNA methylation/hydroxymethylation, and cfDNA fragmentomics in early and multi-cancer detection. Methods: A PRISMA 2020-aligned systematic review was conducted using structured electronic searches and citation chaining for human studies evaluating blood-based cfDNA approaches in localized, preclinical, screening-detected, or otherwise potentially curable cancer. Studies restricted to treatment monitoring, minimal residual disease, or metastatic disease without an early-detection component were excluded. Fourteen primary studies were included. Study quality was assessed using QUADAS-2. Owing to heterogeneity in tumor types, assay platforms, populations, and thresholds, results were synthesized qualitatively. Results: Mutation-based ctDNA assays established the feasibility of detecting early-stage malignancy but were limited by low tumor fraction, genomic heterogeneity, sequencing error, and clonal hematopoiesis. Methylation based assays provided broader tumor-derived signal, very high specificity, and tissue-of-origin information, although stage I sensitivity remained substantially lower than later-stage sensitivity. Fragmentomic approaches based on genome-wide fragment-size distributions, coverage, end motifs, and nucleosome-related patterns showed strong diagnostic performance using low-coverage whole-genome sequencing and were complementary to mutation analysis. Prospective studies demonstrated that cfDNA-based multi-cancer early detection can be integrated into clinical pathways, but positive tests may trigger substantial diagnostic evaluation. QUADAS-2 assessment identified patient-selection bias as the most frequent methodological concern. Conclusion: cfDNA-based early cancer detection is progressing from single-analyte mutation testing toward integrated genomic, epigenomic, and fragmentomic profiling. Methylation and fragmentomics are particularly promising for low-burden disease, while multi-omic approaches may provide the best balance of sensitivity and specificity. Large prospective population studies demonstrating stage shift, acceptable diagnostic harms, cost effectiveness, and ultimately reduced cancer mortality are still required.