Related Experiment Video
Updated: Oct 27, 2025
Predictive Immune Modeling of Solid Tumors
Published on: February 25, 2020
DNA methylation-based prediction of response to immune checkpoint inhibition in metastatic melanoma
Katharina Filipski1,2,3, Michael Scherer4,5,6, Kim N Zeiner7
1Neurological Institute (Edinger Institute), University Hospital, Frankfurt am Main, Germany.
Background:
Therapies based on targeting immune checkpoints have revolutionized the treatment of metastatic melanoma in recent years. Still, biomarkers predicting long-term therapy responses are lacking.
Methods:
A novel approach of reference-free deconvolution of large-scale DNA methylation data enabled us to develop a machine learning classifier based on CpG sites, specific for latent methylation components (LMC), that allowed for patient allocation to prognostic clusters. DNA methylation data were processed using reference-free analyses (MeDeCom) and reference-based computational tumor deconvolution (MethylCIBERSORT, LUMP).
Results:
We provide evidence that DNA methylation signatures of tumor tissue from cutaneous metastases are predictive for therapy response to immune checkpoint inhibition in patients with stage IV metastatic melanoma.
Conclusions:
These results demonstrate that LMC-based segregation of large-scale DNA methylation data is a promising tool for classifier development and treatment response estimation in cancer patients under targeted immunotherapy.
Insights
New DNA methylation signatures can predict how well patients with metastatic melanoma respond to immune checkpoint therapies. This advance offers a promising tool for personalized cancer treatment strategies.
Area of Science:
- Oncology
- Genomics
- Immunotherapy
Background:
- Immune checkpoint inhibitors have transformed metastatic melanoma treatment.
- Biomarkers for predicting long-term responses to these therapies are currently lacking.
Purpose of the Study:
- To identify novel biomarkers for predicting treatment response in metastatic melanoma.
- To develop a machine learning classifier for patient stratification based on DNA methylation data.
Main Methods:
- Utilized reference-free deconvolution of large-scale DNA methylation data (MeDeCom).
- Developed a machine learning classifier based on CpG sites and latent methylation components (LMC).
- Processed DNA methylation data using both reference-free and reference-based computational deconvolution methods (MethylCIBERSORT, LUMP).
Main Results:
- DNA methylation signatures from cutaneous metastases predict therapy response to immune checkpoint inhibition.
- Successfully allocated patients into prognostic clusters using the LMC-based classifier.
Conclusions:
- Latent methylation components (LMC)-based segregation of DNA methylation data is a promising tool.
- This approach aids in classifier development and treatment response estimation for cancer patients receiving immunotherapy.
