DNA Methylation Signature Scoring for Reliable HCC Diagnosis

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing epigenome-wide association studies (EWAS) for diagnosing hepatocellular carcinoma (HCC) face challenges with bias and spurious associations, particularly in small sample sizes, limiting their diagnostic accuracy.

Innovation Solution

A smoothing method for calculating and visualizing data from EWASs to identify highly accurate epigenomic signatures, using a correlation-based approach to derive and validate an epigenome score for HCC diagnosis.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If epigenome-wide association studies (EWAS) are used to diagnose hepatocellular carcinoma, then diagnostic capability is provided, but bias and spurious associations occur particularly in small sample sizes

Engineering Contradiction:
Improvediagnostic accuracyVSAvoidstatistical certainty
Core Design Contradiction:
ReliabilityVSMeasurement precision

Solution Approach 1:

The patent combines results from three independent EWAS studies into a meta-analysis, pooling data to increase statistical power and reduce the impact of small sample sizes. This merging of multiple studies allows for more reliable identification of epigenomic signatures associated with HCC while minimizing spurious associations that would occur in individual small-scale studies.

Inventive Principle:
Principle #5Merging (Combining)

Solution Approach 2:

The patent employs a systematic approach where initial EWAS results are used to generate hypotheses, which are then tested and validated through additional independent studies. This iterative feedback process refines the identification of true epigenomic signatures by confirming findings across multiple cohorts, thereby reducing bias and improving diagnostic accuracy.

Inventive Principle:
Principle #23Feedback

2Quantity of substance

If traditional EWAS methods are used, then genome-wide methylation analysis is performed, but spurious associations are identified due to bias

Engineering Contradiction:
Improvenumber of CpG sites analyzedVSAvoidstatistical certainty
Core Design Contradiction:
Quantity of substanceVSReliability

Solution Approach 1:

The patent transitions from analyzing individual CpG sites in isolation to examining coordinated methylation changes across multiple CpG sites within specific genomic regions. By focusing on local patterns of methylation coordination rather than individual sites, the method identifies more reliable epigenomic signatures while reducing false positives from random variations at single sites.

Inventive Principle:
Principle #3Local quality

Solution Approach 2:

The patent creates a composite epigenomic signature by integrating methylation data from multiple CpG sites and combining results across three independent studies. This composite approach synthesizes information from numerous individual measurements into a unified diagnostic marker set, improving statistical certainty and reducing the impact of random noise or bias in any single measurement.

Inventive Principle:
Principle #40Composite materials

Data Source

PatentEP4419717B1DNA methylation signature for diagnosing hepatocellular carcinoma
Publication Date: 2025.12.03 INST NAT DE LA SANTE & DE LA RECHERCHE MEDICALE (INSERM)
  • EP4419717B1 patent drawingFigure 1
  • EP4419717B1 patent drawingFigure 2
  • EP4419717B1 patent drawingFigure 3

AI summary

The inventors report in the present application the results of applying this innovative analytical approach on three independent EWASs for deriving and validating an epigenome score that exhibited a high diagnostic accuracy for detecting hepatocellular carcinoma (HCC). Indeed, they have performed in silico replication studies on two independent DNA methylome datasets using the 105 CpG probes of the HCC Epigenome Score. All the 13 loci were significantly associated with the HCC phenotype in both replication studies. Consistently, the random effect meta-analysis confirmed the significance of the 13 top loci. The inventors performed dose-response probit regression analysis to assess the association between the HCC Epigenome Score using the 586 samples. The HCC Epigenome score was significantly associated with the risk of HCC with a gradual increase in the risk of HCC according to the increase in the HCC Epigenome Score. Accordingly, the present invention relates to the use of a DNA methylation signature for diagnosing hepatocellular carcinoma in a subject.