Abstract
Accurate risk assessment of local recurrences or metastases is essential for melanoma treatment, but current clinical parameters are suboptimal. Therefore, we conducted a comprehensive DNA methylation analysis to identify prognostic markers that could improve risk prediction in patients with melanoma. We integrated methyl-binding domain sequencing, RNA sequencing, Infinium HumanMethylation450 analyses, and The Cancer Genome Atlas data to identify potential prognostic DNA methylation markers. Validation was performed using methylation-specific PCR in 2 independent melanoma cohorts as well as an in silico validation using The Cancer Genome Atlas data. Cox proportional hazards models and backward stepwise elimination determined the prognostic value of candidate markers. LY75 promoter methylation, Breslow thickness, and metastatic disease at diagnosis were significant independent predictors of recurrence-free survival. LY75 promoter methylation was also found to be an independent predictor of metastatic disease development in patients with localized stage I/II melanoma specifically. This association was confirmed in a second validation cohort and The Cancer Genome Atlas melanoma dataset. LY75 promoter methylation identifies patients at high risk of recurrences or metastases, independent of conventional prognostic parameters. This marker could help refine patient selection for sentinel node biopsies, imaging, (neo)adjuvant treatment, and closer monitoring, ultimately leading to improved clinical outcomes.
| Original language | English |
|---|---|
| Pages (from-to) | 455-469.e5 |
| Number of pages | 20 |
| Journal | Journal of Investigative Dermatology |
| Volume | 146 |
| Issue number | 2 |
| Early online date | 2025 |
| DOIs | |
| Publication status | Published - Feb 2026 |
Keywords
- Epigenetics
- LY75
- Melanoma
- Prognostic biomarker
- promoter DNA methylation
Fingerprint
Dive into the research topics of 'Comprehensive Analysis of the Melanoma DNA Methylome Identifies LY75 Methylation as an Independent Marker Predicting Poor Clinical Outcome'. Together they form a unique fingerprint.Cite this
- APA
- Author
- BIBTEX
- Harvard
- Standard
- RIS
- Vancouver