Liquid biopsy has emerged as a minimally invasive approach for the molecular characterization and longitudinal monitoring of primary central nervous system (CNS) tumors. Although extensively validated in systemic malignancies, its clinical application in CNS tumors is challenged by the blood–brain barrier, low analyte abundance, and heterogeneous assay performance. Recent advances have expanded the spectrum of detectable tumor-derived components, including circulating tumor DNA (ctDNA), cell-free DNA, extracellular vesicles, RNA species, nucleosomes, metabolites, and lipids, across multiple biofluids such as cerebrospinal fluid (CSF), plasma, serum, urine, saliva, and tears. The aim of this study was to review the biological foundations, analytes, biofluids, clinical applications, and technical limitations of liquid biopsy in primary CNS tumors, with emphasis on diagnostic, prognostic, and surveillance value. We synthesized current evidence on tumor-derived analytes detectable through liquid biopsy, their molecular correlates, and their performance across biofluids. CSF is consistently the most informative biofluid for CNS tumors, with higher analyte concentration and superior concordance with tumor tissue compared with plasma. CtDNA in CSF reliably identifies hallmark alterations, including IDH1/2, H3K27M, TERT, BRAF, ATRX, TP53, 1p/19q codeletion, and MYCN amplification, thereby enabling better diagnosis, molecular classification, and therapeutic stratification. Plasma-based assays are less sensitive but remain valuable for longitudinal monitoring, especially when combined with ultrasensitive sequencing or fragmentomic approaches. Emerging biomarkers, including nucleosome footprints, exosomes, proteins, microRNA (miRNA)/long noncoding RNA (lncRNA)/circular RNA (circRNA) signatures, lipidomic panels, and metabolites such as d-2-hydroxyglutarate, show potential for integration into multimodal diagnostics. Liquid biopsy provides a powerful and rapidly evolving tool for the minimally invasive molecular assessment of CNS tumors. While CSF remains the optimal matrix for diagnosis and characterization, advances in ultrasensitive detection methods increasingly support the feasibility of plasma and urine for longitudinal follow-up. Integrating liquid biopsy with advanced imaging and tissue-based data will likely transform diagnostic accuracy, therapeutic decision-making, and real-time monitoring of CNS tumors.
Epilepsy is a complex neurological disorder characterized by pathological processes that unfold across multiple biological scales, from cellular excitability and synaptic integration to large-scale network dynamics observable in electroencephalographic (EEG) recordings. While traditional analytical approaches have provided valuable insights, they often fail to capture high-dimensional and nonlinear structure of contemporary electrophysiological and clinical datasets. Consequently, machine learning (ML) has emerged as a powerful analytical framework in epilepsy research, although its rapid adoption has revealed a growing gap between algorithmic performance and biological interpretability. This review examines ML methods operating across three analytically distinct yet interconnected levels: (i) unsupervised learning for cellular-level phenotyping using high-dimensional electrophysiological data; (ii) supervised learning for EEG-based seizure detection and prediction; and (iii) multiscale modeling frameworks integrating neuronal and network dynamics. Rather than providing an exhaustive catalog of algorithms, we focus on inferential assumptions underlying ML applications, the methodological pitfalls constraining generalization and clinical relevance, and how ML-derived representations can be interpreted within established neurophysiological theory. We highlight that unsupervised ML facilitates identification of latent excitability phenotypes and trajectories obscured in traditional univariate analyses, while supervised ML has substantially advanced automated seizure detection and prediction, despite persistent challenges related to data leakage, class imbalance, and ambiguous preictal labeling. We argue that the most promising direction lies in embedding ML within multiscale mechanistic models, where data-driven inference facilitates parameter estimation and hypothesis generation rather than black-box prediction. By prioritizing interpretability, rigorous validation, and cross-scale integration, ML-enhanced multiscale frameworks offer a path toward clinically actionable models of epilepsy.
Purkinje neurons (PNs), the sole output neurons of the cerebellar cortex, control motor activity, cortical excitability, and seizure propagation. They regulate cerebello-thalamo-cortical circuits by inhibiting deep cerebellar nuclei, which are increasingly linked to epilepsy. While epilepsy-related alterations in cortical and hippocampal neurons are well documented, cerebellar PNs remain understudied. Disruption of PN-mediated inhibition may lead to a breakdown in network regulation and promote seizure activity. This study investigated whether prolonged epileptic activity alters both passive and active electrophysiological properties of PNs and whether multivariate analysis can reveal biophysical abnormalities associated with epilepsy. Additionally, we explored whether spectral EEG dynamics reflect or predict these cellular changes. Whole-cell patch-clamp recordings were obtained from Crus II PNs in amygdala-kindled rats to extract 12 intrinsic membrane properties. Multivariate analysis, utilizing principal component analysis (PCA) and K-means clustering, identified latent electrophysiological subtypes. Concurrent EEG signals from the hippocampus and amygdala were analyzed using Fast Fourier Transformation (FFT), PCA, and supervised classification to track seizure-related spectral dynamics. Results revealed that chronic seizures reduce PN excitability and induce a convergence toward specific biophysical states. EEG analysis uncovered latent spectral patterns associated with behavioral seizure severity. A Random Forest Classifier trained on spectral band power predicted Racine stages reasonably, highlighting Beta and Gamma bands being particularly influential. These findings demonstrate that chronic seizures drive coordinated changes in the properties of cerebellar neurons and EEG dynamics. This multiscale approach reveals consistent electrophysiological changes at cellular and network levels, underscoring the cerebellum’s role in epileptogenesis and supporting its potential as a therapeutic target.
The relationship between epilepsy, obesity, and metabolic syndrome (MetS) has emerged as a rapidly evolving area of neurobiology inquiry. Emerging evidence suggests that epilepsy extends beyond neuronal hyperexcitability, reframing it as a systemic condition characterized by significant metabolic dysregulation. Converging supports a bidirectional relationship while seizures, antiseizure medications (ASM), and neuroinflammation induce exacerbate potentiate epileptogenesis through shared molecular pathways. At the cellular level, chronic epileptic activity induces oxidative stress, mitochondrial dysfunction, and the activation of microglia and astrocytes. This, in turn, leads to the release of pro-inflammatory cytokines including TNF-α, IL-1β, and IL-6. These mediators traverse the blood-brain barrier (BBB), subsequently modifying insulin signaling, and disrupting glucose homeostasis, which collectively fosters a pro-inflammatory and insulin-resistant environment. Furthermore, antiseizure medications such as valproate can exacerbate these effects by directly impairing insulin receptor signaling and altering adipokine production, ultimately contributing to weight gain and systemic metabolic dysregulation. Obesity and MetS induce neuroinflammatory and excitotoxic states that promote seizure onset via leptin resistance, reduced adiponectin levels, and compromised AMP-activated protein kinase (AMPK) signaling. Emerging evidence emphasizes the gut-brain axis as a crucial regulator in this reciprocal interaction. Dysbiosis, altered microbial metabolites (e.g., short-chain fatty acids), and heightened intestinal permeability facilitate systemic inflammation and BBB disruption, enhancing neuronal excitability. Insulin resistance in the brain disrupts synaptic transmission, impairs mitochondrial biogenesis, and compromises redox equilibrium, perpetuating a pathological cycle linking metabolic stress to epileptic activity. This review synthesizes the cellular, molecular, and systemic pathways connecting epilepsy, obesity, and MetS, and proposes that epilepsy be reconceptualized as a neuro-metabolic disorder. Insights into these convergent pathways provide a rationale for novel therapeutic strategies that simultaneously target seizure control and metabolic regulation, encompassing microbiota modulation, antioxidant therapy, and insulin-sensitizing interventions with the overarching aim of restoring neuro-metabolic homeostasis.
