Oncology outcomes
DNA methylation signatures separate HCC from nearby liver
Open access · cc by · source: Europe PMC
Promoter methylation arrays in 30 HCC tumours vs surrounding tissue produced a validated signature that distinguished HCC from adjacent liver (independent of major risk factors), with other methylation panels tracking risk factors, progression, and survival after therapy.
Study at a glance
- Design
- Other — Promoter methylation bead-array discovery in paired tumour/surrounding tissue with independent validation series
- N
- N=30 · Discovery: 30 HCC tumours with matched surrounding tissue (38 patients selected; 30 with paired samples)
- Population
- Surgical HCC patients at Edouard Herriot Hospital (Lyon) with cryopreserved tumour tissue
- Outcome
- Methylation signatures distinguishing HCC and correlating with risk factors/survival
Structured fields used in claim comparison tables when every cited study has a complete layer.
Key findings
They developed and validated a methylation signature that distinguished HCC from surrounding tissue and from other tumour types, independent of major risk factors. Distinct methylation panels associated with risk factors and progression; an independent promoter panel strongly correlated with survival after cancer therapy. A 20-CpG predictor heat map was shown on an independent series.
Methodology
At Edouard Herriot Hospital (Lyon), authors profiled promoter methylation with Illumina bead arrays covering 1505 CpG sites in 807 cancer-related genes across 30 HCC tumours and matched surrounding tissue, selected classifying CpG sites, and validated signatures in an independent HCC series.
Limitations
A methylation signature that classifies tissue is not by itself a population screening blood test. Survival correlations need careful external validation before treatment decisions. Array coverage is a cancer-gene panel, not the entire methylome.
How this study connects
Role on claims
Each row is a claim on a concept or method page where this paper supports, challenges, or qualifies the statement. Roles are hand-checked — not a model guess.
Higher tumour EZH2 mRNA associates with portal-vein invasion after HCC resection.
In 66 surgical HCC tumours, EZH2 mRNA exceeded matched nontumour liver and high expression tracked portal-vein invasion (79% vs 39%) without a significant disease-free survival split — a clinicopathological progression signal, not a proven survival assay in this cohort.
Scope note — different molecular layer — promoter methylation signatures, not EZH2 mRNA invasion associations
Limits the claim's scope: a different population, assay, or outcome.
Validated DNA methylation signatures can separate HCC from surrounding liver.
Illumina promoter arrays in 30 paired HCC/surrounding samples produced a validated methylation signature that distinguished tumour from adjacent liver independent of major risk factors, with other panels linked to progression and survival after therapy.
Evidence for the claim as stated.
EZH2 marks invasion risk in resected human tumours; methylation signatures classify tissue epigenetically; miR-122 exosomes test a delivery/chemosensitisation idea in models. Shared HCC context does not make them one biomarker or one therapy.
Evidence for the claim as stated.
EZH2’s invasion association without a DFS split is a different clinical claim from methylation panels that correlate with survival after therapy — both can be true without interchangeable cutoffs.
Evidence for the claim as stated.
Open questions
Tensions this paper is part of
From concept pages' “where studies disagree.” Disagreement means the same question; scope means different assays, populations, or outcomes.
EZH2 marks invasion risk in resected human tumours; methylation signatures classify tissue epigenetically; miR-122 exosomes test a delivery/chemosensitisation idea in models. Shared HCC context does not make them one biomarker or one therapy.
EZH2’s invasion association without a DFS split is a different clinical claim from methylation panels that correlate with survival after therapy — both can be true without interchangeable cutoffs.
Related papers in this topic
Same topic cluster — not a recommendation engine.
- miR-122 exosomes from fat stem cells sensitize HCC models to chemo
- A 22-gene classifier flags early metastasis after prostatectomy
- HDAC2 IHC tracks shorter PSA relapse in prostate cancer
- Five integrated genomic subgroups track prostate-cancer relapse risk
- Higher tumour EZH2 tracks portal-vein invasion in HCC