Background DNA methylation abnormalities are pervasive in pituitary neuroendocrine tumors (PitNETs). The feasibility to detect methylome alterations in circulating cell-free DNA (cfDNA) has been reported for several central nervous system tumors but not across PitNETs. The aim of the study was to use the liquid biopsy approach to detect PitNET-specific methylation signatures to differentiate these tumors from other sellar diseases. Method We profiled the cfDNA methylome (EPIC array) of 59 serum and 41 plasma liquid biopsy specimens from patients with PitNETs and other CNS diseases (sellar tumors and other pituitary non-neoplastic diseases, lower-grade gliomas and skull base meningiomas) or nontumor conditions, grouped as non-PitNET. Results Our results indicated that, despite quantitative and qualitative differences between serum and plasma cfDNA composition, both sources of liquid biopsy showed that patients with PitNETs presented a distinct methylome landscape compared to non-PitNETs. In addition, liquid biopsy methylomes captured epigenetic features reported in PitNET tissue and provided information about cell type composition. Using liquid biopsy-derived PitNETs-specific signatures as input to develop machine-learning predictive models, we generated scores which distinguished PitNETs from non-PitNETs conditions, including sellar tumor and non-neoplastic pituitary diseases, with accuracies above ~93% in independent cohort sets. Conclusions Our results underpin the potential application of methylation-based liquid biopsy profiling as a noninvasive approach to identify clinically relevant epigenetic markers to diagnose and potentially impact the prognostication and management of patients with PitNETs.

In this review, we summarize the current approaches used to detect glioma tissue-derived DNA methylation markers in liquid biopsy specimens with the aim to diagnose, prognosticate and potentially track treatment response and evolution of patients with gliomas.

Background Detection of distinct epigenetic biomarkers in circulating cell-free DNA (cfDNA) of liquid biopsy (LB) specimens (e.g. blood) fosters opportunity for prognostication of central nervous system (CNS) tumors and has not been thoroughly explored in patients with meningiomas. Material and Methods We profiled the cfDNA methylome (EPIC array) in serum specimens from patients with meningiomas (MNG; n= 63) and harnessed internal and external meningioma tissue methylome data with reported follow up (n=48). To predict recurrence risk (RR), we consolidated a tissue cohort with at least 5 years of follow up and divided them into confirmed recurrence (CR; either reported progressive disease in post-surgical imaging, or additional resections following initial surgery) and confirmed no-recurrence (CNR: no confirmed disease progression w/in at least 5-years of follow-up). Then through application of an iterative process consisting of multiple tissue- and serum-based supervised analyses, we identified risk-specific methylation markers with serum specific features which, when inputted into a random forest algorithm allowed for segregation of both tumor tissue and liquid biopsy specimens according to recurrence risk. We estimated immune cell composition using MethylCIBERSORT, where a reference methylome atlas of chosen immune cell types was utilized to deconvolute the MNG samples. Results The resulting recurrence risk classifier demonstrated an appreciable predictive power in classifying samples as high or low recurrence risk across the tumor tissue cohort (ACC: 87.5%, CUI+: 85.2%). When compared to another classifier, our model demonstrated statistically significant agreement across primary meningioma samples (κ=0.269, p=0.002), and more accurately predicted samples to recur across an expanded time window (time to recurrence >5yrs). Across resulting liquid biopsy classifications, recurrence risk subgroups were analogous with reported risk factors, including WHO grade, extent of resection, and tumor location. Recurrence risk subgroups (high and low) also demonstrated differential estimated immune cell contributions, with low-risk samples exhibiting a “hot” profile, or enrichment of B-Cells, CD56- and CD4 T-Cells, and natural killer cells. Notably, the estimated neutrophil to lymphocyte ratio, previously purported to be relevant to tumor prognosis, was appreciably higher for those meningioma samples with the highest recurrence risk. Conclusion DNA methylation markers identified in the serum are suitable for the development of machine learning-based models which present high predictive power to prognosticate patients with meningioma and estimate a differential immune profile across recurrence risk groups. After validation in an external cohort, this noninvasive approach may improve the presurgical therapeutic management of patients with meningiomas.

Background Tumor-infiltrating immune cell compositions have been previously correlated to encouragement or inhibition of tumor growth. This association highlights immune-landscape profiling through non-invasive methods as a crucial step in approaches to treatment of patients with meningioma (MNG), a prevalent primary intracranial tumor. Genome-wide DNA methylation patterns can aid in definition and assessment of cell compositions in liquid biopsy serum specimens, and allow for development of machine-learning models with predictive capabilities. Methods We profiled the cfDNA methylome (EPIC array) in liquid biopsy specimens from patients with MNG (n = 63) and nontumor controls (n = 6). We conducted both unsupervised epigenome-wide and supervised analyses of the meningioma methylome. Estimation of immune cell composition was conducted using Python-based methodology, where a reference methylome atlas of chosen cell types (B-cells, CD4- and CD8T-cells, neutrophils, natural killer cells, monocytes, cortical neuron, vascular endothelial cells, and healthy meninge) was used to deconvolute the MNG samples. Recurrence risk was estimated using an existing methylation-based Random-Forest classifier previously reported and validated, adapted to our serum-based cohort through employment of translatable meningioma subgroup-specific methylation markers (differentially methylated probes). Results We identified four distinct genome-wide methylation subgroups (k-clusters) of MNG which presented differential tumor micro-environments across all cell types investigated. Application of the DNA methylation-based Random-Forest classifier allowed for categorization of primary MNG serum samples into estimated recurrence-risk subgroups. Significantly contrasting micro-environments for the subgroups were observed across several cell-types, with those MNG more likely to recur displaying depletion in cell types reported to improve anti-tumoral response in many tumors (e.g. T-Cells). Conclusions DNA methylation based deconvolution allowed for detection of contrasting tumor microenvironment compositions across MNG methylation subtypes and recurrence-risk estimation subgroups. These results suggest that microenvironment profiling can be informative of probable tumor behavior and prognostic outcomes, helping guide therapeutic approaches towards treatment of patients with MNG.

Background DNA methylation abnormalities are pervasive in pituitary neuroendocrine tumors (PitNETs). The feasibility to detect these molecular alterations in circulating cell-free DNA (cfDNA) has been reported for several central nervous system tumors but not across PitNETs. Hypothesis PitNET-specific methylation signatures detected in liquid biopsy specimens differentiate PitNETs from other sellar diseases. Method We profiled the cfDNA methylome (EPIC array) of 44 serum and 34 plasma liquid biopsy (LB) specimens from patients with PitNETs and other CNS (craniopharyngiomas, other pituitary diseases, gliomas, meningiomas) or nontumor conditions, grouped as non-PitNET. Results Our results indicated that, despite quantitative and qualitative differences between serum and plasma cfDNA composition, both sources of LB showed that patients with PitNETs presented a distinct methylome landscape compared to non-PitNETs. In addition, LB methylome captured epigenetic features reported in PitNET tissue. Using LB-derived PitNETs-specific signatures as input into a machine-learning algorithm, we generated a score that distinguished PitNETs from other pituitary and CNS diseases with high accuracy in an independent set. Conclusions Our results underpin the potential application of a methylation-based LB as a noninvasive approach to identify clinically relevant epigenetic markers to diagnose and potentially impact the management of patients with PitNETs. Presentation: No date and time listed