LNA Blocker-Enhanced Multiplex Chip-Based Digital PCR for Ultrasensitive Quantification of Liver-Specific DNA Methylation in Plasma

Abstract Sensitive and selective quantification of rare methylated DNA in plasma remains a major analytical challenge. For methylation targets located in CpG-sparse regions, bisulfite conversion produces only subtle sequence differences between methylated and unmethylated alleles, thereby hindering selective amplification. Here, we report a blocker-enhanced multiplex chip-based digital PCR (cdPCR) strategy for the absolute quantification of liver-specific DNA methylation in plasma. The assay incorporates Locked Nucleic Acid (LNA)-modified blockers to selectively suppress unmethylated background templates, thereby enhancing single-base mismatch discrimination after bisulfite conversion. Seven liver-specific methylation sites within TBX15 were selected and organized into two quadruplex cdPCR panels together with an internal reference gene, enabling multiplex partition-level quantification. The platform achieved a limit of detection of 1 copy/μL, a limit of quantification of 0.2% methylation fraction, and excellent linearity over 1−10,000 copies/μL. In the primary cohort and independent external cohorts comprising 815 plasma samples, the resulting methylation profiles were used to develop diagnostic models using extreme gradient boosting (XGBoost) and logistic regression. The XGBoost model distinguished hepatocellular carcinoma from cirrhosis with an AUC of 0.881 in the test set and retained strong performance in early-stage and seronegative disease, outperforming alpha-fetoprotein and des-gamma-carboxy prothrombin. Additional disease-control and pan-cancer cohorts supported the specificity of the methylation-based models. These results demonstrate that coupling LNA-mediated background suppression with chip-based digital partitioning enables selective and sensitive plasma methylation analysis and provides a practical analytical framework for low-abundance nucleic acid detection.