Differential immune-cell landscape in tissue and liquid biopsy specimens is associated with recurrence risk across meningioma methylation-subtypes

Document Type

Conference Proceeding

Publication Date


Publication Title

Cancer Res


Background. Tumor-infiltrating immune cells can favor or inhibit tumor growth, constituting promising and alternative targets for treatment of patients with meningiomas (MNG), a prevalent primary intracranial tumor with no known therapy beyond surgery and radiation. Immune-cell composition can predict therapy response in other diseases, therefore, profiling immune-cell infiltrates in these tumors is crucial, preferentially by non-invasive approaches. Genome-wide methylation arrays are powerful tools to classify and assess cell composition either in tumor tissue or with liquid biopsy. Objective: We aim to apply methylation-based classifiers to estimate immune-cell composition of tumor tissue and serum specimens to predict recurrence-free survival in patients with MNG. Methods: We conducted unsupervised and supervised analyses of the methylome of a multicenter MNG cohort composed of 326 tissue, 60 serum and 9 control specimens. Immune-cell composition (B-cells, CD4-T & CD8-T cells, neutrophils, natural killer cells, and monocytes) and recurrence risk were estimated using methylation-based classifiers. We used the point-biserial correlation method to analyze associations across categorical clinical and molecular features. Results: We identified four distinct methylation subgroups of MNG (in serum or tissue) that presented a differential immune cell landscape composition. In the tissue, MNG methylation subgroups were mutually exclusive in enrichment or depletion of NK-, cytotoxic T- or monocyte-cell methylation signatures and in the serum, they were enriched of monocyte- and B-cell signatures. Distinct Immune-cell signatures (Monocytes, CD8T and Neutrophils) were correlated with risk of recurrence or grade in MNG tissue and/or liquid biopsy specimens. Conclusion: DNA methylation-based deconvolution allowed for the detection of differential immune-cell composition across MNG methylation-subtypes in tissue that could be detected in liquid biopsy specimens. Distinct immune-cell signatures correlated with MNG recurrence risk/grade (e.g. T-cell) in our cohort were reported to improve antitumoral immune response in many tumors. These results suggest that profiling tumor immune-cell landscape can inform tumor behavior and provide insight in application of existing and novel therapeutic approaches in MNG.

PubMed ID

Not assigned.