Neuro-Oncology Advances

BackgroundPaediatric and adolescent gliomas and glioneuronal tumours remain challenging to assess non-invasively. Amide proton transfer (APT) chemical exchange saturation transfer (CEST) MRI has shown promise in adult gliomas but has not been well studied in younger patients. PurposeTo assess whether APT CEST signal can act as a non-invasive surrogate of tumour proliferation in adolescent CNS tumours by correlating it with 18F-choline PET uptake (SUV) as a proxy for membrane synthesis / proliferative activity. MethodsTen adolescent patients (14-19 yrs) with confirmed or suspected gliomas / glioneuronal tumours underwent simultaneous APT CEST and 18F-choline PET-MRI. Regions of interest (ROIs) corresponding to non-enhancing, enhancing, necrotic tumour, and contralateral white matter were delineated. Mean APT signal intensity (SI) and PET SUV were extracted per ROI. Nonparametric statistics and Spearmans correlation analyses were performed. ResultsAPT SI was significantly elevated in enhancing, non-enhancing, and necrotic tumour ROIs compared to normal white matter (p<0.001). 18F-choline SUV was elevated in enhancing and necrotic ROIs vs white matter, but not significantly so for non-enhancing tumour (p=0.02). A strong correlation between whole-tumour APT SI and 18F-choline SUV was seen (Spearman {rho}=0.86, p<0.001). ConclusionOur results indicate that APT CEST is feasible in adolescents and may reflect proliferative tumour burden. The detection of elevated APT SI even in non-enhancing tumour regions suggests potential utility in monitoring non-contrast-enhancing disease. Larger cohorts and multimodal correlation (e.g. Ki-67, amino acid PET) are warranted to confirm and extend these findings.

PurposeTo externally evaluate three binary classification models designed to differentiate the molecular subtype of pediatric low-grade glioma (pLGG) between BRAF Fusion, BRAF Mutation, and Wild Type on T2-weighted magnetic resonance imaging using self-supervised transfer learning, which enables effective performance in a low data setting. Materials and methodsThis retrospective study evaluates pLGG molecular subtyping models, pre-trained using data collected at Dana Farber Cancer Institute/Bostons Childrens Hospital, on two datasets from the Hospital for Sick Children, one consisting of patients identified from the electronic health record between January 2000 to December 2018 (n=336) and another consisting of patients identified from the electronic health record between January 2019 to April 2023 (n=87). These datasets consist of T2-weighted MRI with pLGG and corresponding genetic marker identifications, labelled as BRAF Fusion, BRAF Mutation, or Wild Type. The datasets included manually annotated ground-truth segmentations that were used in the classification pipeline during evaluation. The models were evaluated using the area under the receiver operating characteristic curve (AUC). To acquire a per-class probabilities across all three considered molecular subtypes, we used the output probabilities from each binary model as logits input to a Softmax function. These probabilities were used to determine the AUC of the models on each evaluated dataset. ResultsThe models performed achieved a macro-average AUC of 0.7671 on the newer dataset from the Hospital for Sick Children but achieved a lower macro-average AUC of 0.6463 on the older dataset from the Hospital for Sick Children. ConclusionsThe evaluated pLGG molecular subtyping models have the potential for effective generalization but may require further fine-tuning for consistent performance across varying datasets.

Brain tumors are among the most lethal cancers with gliomas representing the most morphologically complex type. Precise and time efficient glioma segmentation and classification are essential for accurate diagnosis, treatment planning, and patient monitoring. Magnetic resonance imaging (MRI) remains the primary imaging modality for noninvasive glioma assessment. This review systematically analyzes deep learning (DL) and artificial intelligence (AI) approaches for brain tumor segmentation and classification. Thirty one studies, out of 310 published between 2022 and 2025, met the inclusion criteria, among which 8 performed both segmentation and classification tasks. For segmentation, most of the studies used publicly available multiparametric MRI datasets. Segmentation performance varied by model and tumor region, with those focused on the whole tumor region achieving the highest Dice Score Coefficient (DSC). Classical U Nets achieved DSC scores around 80%, while advanced models integrating residual or attention modules exceeded 90%. Two main classification tasks were performed: tumor type and glioma staging. Classification models primarily relied on learned features extracted from multiparametric MRI using DL models, reporting an accuracy from 91.3% to 99.4%, with sensitivity and specificity typically above 95%, indicating robust predictive performance. Surprisingly, explainable AI approaches were infrequently applied, highlighting the persistent need for greater model transparency to foster clinical trust. Overall, these results demonstrate the strong potential of current AI based segmentation and classification pipelines. These methods can help clinicians accelerate the decision making process, increasing both the accuracy and efficiency of brain tumor diagnosis. These approaches may also support the development of personalized treatment plans tailored to each patient.

BackgroundGlioblastoma (GBM), Isocitrate dehydrogenase-wildtype (IDH-wt) is characterised by diffuse infiltration, with progression often arising from perilesional tissue and occult white-matter damage. We investigated whether radiomics from the T2/FLAIR-defined oedema and the structural disconnectome improve prediction of progression-free survival (PFS). MethodsWe retrospectively analysed 387 adults with newly diagnosed GBM, IDH-wt treated at a single tertiary centre (2005-2020). A deep-learning pipeline segmented enhancing tumour, non-enhancing tumour, and oedema on pre-operative MRI; lesion masks were propagated to normative tractography to derive disconnectome maps. 3-D shape radiomic features extracted for each segmented region underwent appropriate feature selection. Finally, 10 tumour and 9 oedema radiomics were combined with 6 clinical features to train 3 survival models (Random Survival Forest (RSF), XGBoost, Cox proportional hazards (CPH)) that were evaluated on a held-out 20% test set using Harrells C-index, Kaplan-Meier risk stratification and time-dependent ROC curves. ResultsThe best performance was achieved by RSF using all clinical and radiomic features (C-index 0.665 vs 0.595 for clinical features only, p=0.088). Models including oedema radiomics outperformed those using tumour radiomics alone, and disconnectome features, derived from both tumour and oedema regions, were repeatedly selected among the top predictors across algorithms. Combining radiomic and clinical features improved risk stratification and 12-month early-versus-late recurrence classification (AUC 0.704 vs 0.582 for clinical features alone). ConclusionsIntegrating perilesional oedema and white-matter disconnectome MR features with clinical and molecular data enhances prediction of PFS in GBM, IDH-wt. These network-aware, multimodal survival models may support personalised risk-adapted treatment strategies pending external validation. Key Points- GBM IDH-wt exhibits a high recurrence rate despite aggressive treatment. - Addition of high-dimensional oedema and disconnectome radiomic features to clinical features showed consistent improvement in the test performance of 3 ML models. - This can support informed clinical decision-making. Importance of the StudyPrediction of progression free survival (PFS) for a patient with highly recurrent glioblastoma IDH-wt traditionally relies on clinical history, demographics, and molecular markers of the tumour. Recent literature reveals the tumours disruptive nature through its invasion of white-matter tracts and identifies its microenvironment, particularly the perilesional oedema, as a harbour of treatment resistant tumour cells. This study is the first to combine high-dimensional radiomic features of the tumour, the oedema, and their disconnectome with clinical and treatment factors to predict PFS. Using 3 model architectures (XGBoost, RSF, and CoxPH), we demonstrate consistent directional improvements in performance, on addition of radiomic features to clinical baseline models. Furthermore, oedema and disconnectome radiomics are identified as top predictor features across algorithms. This proof-of-concept study provides a reproducible multimodal pipeline, reaffirms the usability of MR radiomics, and identifies features of the oedema and the structural connectome as promising biomarkers, demanding large-scale external validation.

BackgroundGlioblastoma multiforme (GBM) is an aggressive brain tumor that is notoriously resistant to treatment, with an average survival time of 17 months. While the overall outcome is poor for both males and females, sex differences in GBM incidence and outcome suggest sex-specific biological mechanisms underlie tumorigenesis. In contrast, low-grade glioma (LGG) is a less aggressive brain tumor that tends to have a better prognosis and a longer survival time. MethodsTo understand mechanisms contributing to treatment resistance in GBM in both males and females, we inferred gene regulatory networks (GRNs) for males and females with LGG and GBM using RNA-seq data from The Cancer Genome Atlas (TCGA). We analyzed these to identify both sex-specific and sex-stratified gene regulation in GBM. ResultsWe found few sex-specific differences in gene regulation in individuals with LGG, consistent with the lack of evidence for significant clinical endpoints dependent on sex. However, in GBM-we found sex-specific differential targeting of several pathways, including hypoxia and related pathways (carbohydrate metabolism, innate immune processes, and extracellular matrix pathways) known to be dysregulated in hypoxic conditions. In comparing between individuals with GBM, we found that females exhibited a greater degree of co-regulation between hypoxia with the aforementioned downstream pathways than did males. ConclusionsOur results suggest that dysregulation of hypoxia-related pathways in GBM plays a female-specific role in resistance to treatment and overall outcomes.

BackgroundPATZ1 fusion-positive central nervous system (CNS) tumors frequently harbor MN1::PATZ1 fusions as driver mutations, provisionally classified as a rare DNA methylation class of low-grade neuroepithelial tumors. Radiographically, they resemble pilocytic astrocytomas with tumor and cystic components, but their supratentorial cortex location and higher recurrence rates are distinguishing features. An intermediate clinical course, despite focal high-grade histopathology, underscores the need for longitudinal molecular and immune analyses to refine classification and standard therapy. Case SummaryA female pediatric patient presented with neurological symptoms, including headache and right upper extremity weakness. MRI revealed a large cystic lesion in the left frontal lobe, leading to a differential diagnosis of low-grade glioma and ependymoma. Genomic analysis identified an MN1::PATZ1 fusion. The tumor recurred after gross total resection prompting a second resection. Transcriptomic and histopathologic assessments identified multiglial lineage, and high-grade features closely related to adult glioblastoma alongside pro-inflammatory activity in the primary tumor. The recurrent tumor showed reduced malignancy, and oligodendroglioma-like features. Increased MHC gene expression, immune checkpoint receptors (PDCD1, CTLA4, TIGIT,TIM3), T cell regulators (CXCR6), and elevated macrophage frequency, coupled with reduced PD-L1 in the recurrent tumor, suggest a complex anti-tumor immune response constrained by T cell dysregulation. This case, along with two other MN1::PATZ1 fusion-positive tumors, identifies a distinct transcriptomic subtype separate from circumscribed astrocytic glioma, highlighting upregulation of growth factor receptor pathways, like PI3K/AKT, and immune dysfunction linked to recurrence. ConclusionLongitudinal multi-omics analyses of recurrent MN1::PATZ1 fusion-positive CNS tumors revealed tumor maturation, immune dysfunction, and potential therapeutic targets. Introductory ParagraphPATZ1 fusion-positive central nervous system (CNS) tumors are rare, predominantly pediatric and frequently recurrent neoplasms provisionally classified as neuroepithelial tumors. Their pronounced histopathological and clinical heterogeneity, along with limited immunological characterization complicates their treatment standardization. We report a new case of an MN1::PATZ1 fusion-positive CNS tumor with recurrence, highlighting its radiographic similarities to low-to-intermediate grade pediatric glioma. Longitudinal multi-omics analyses of this case, along with additional MN1::PATZ1 fusion-positive CNS tumors, however, delineates a transcriptome subtype resembling adult high-grade glioma, with activated oncogenic and pro-inflammatory programs. The recurrent tumor exhibits features of decreased malignancy and enhanced glial differentiation, phenotypically shifting towards oligodendroglioma, suggesting tumor maturation. This was accompanied by increased antigen presentation programs, indicating immune engagement, while increased immune checkpoint expression and microglia/macrophage frequency indicate T cell exhaustion and immunomodulation, respectively. This longitudinal study highlights potential therapeutic strategies targeting both the tumor and its immune environment in MN1::PATZ1 fusion-positive CNS tumors.

BackgroundIsocitrate dehydrogenase wildtype (IDHwt) histologic grade 2 adult diffuse gliomas represent a highly heterogeneous entity, and the effects of postoperative adjuvant temozolomide (TMZ) therapy, as well as the predictive value of chemotherapy-related biomarker O6-methylguanine-DNA methyltransferase promoter (MGMT-p) methylation, remain to be further investigated. MethodsA Discovery dataset, comprising 108 IDHwt histologic grade 2 gliomas obtained from three public resources, was constructed to investigate the impact of TMZ on patient survival. Furthermore, an independent Validation dataset, consisting of 123 IDHwt grade 2 gliomas, was created to validate the effect of TMZ on patient survival. Kaplan-Meier overall survival (OS) analyses and Cox proportional hazard models were used. ResultsMultivariable analysis in the validation dataset demonstrated that temozolomide (TMZ) chemotherapy was an adverse independent prognostic factor for survival in patients with histologic grade 2 IDH-wildtype (IDHwt) gliomas (HR = 2.19, P = 0.033). Subgroup analyses further revealed that the detrimental effect of TMZ was mainly confined to MGMT-p unmethylated tumors (Discovery cohort: TMZ vs noTMZ, median overall survival [OS]: 37.0 vs 130.0 months, log-rank P = 0.005, HR = 1.89; Validation cohort: TMZ vs noTMZ, median OS: 34.4 vs >120 months, log-rank P = 0.004, HR = 3.39). Moreover, in the validation cohort, TMZ therapy remained associated with significantly poorer survival in the NEC (Not Elsewhere Classified) subgroup (TMZ vs noTMZ, median OS: 73 vs >120 months, P = 0.017), while in molecular GBMs it did not reach statistical significance but still showed a trend toward worse survival (TMZ vs noTMZ, median OS: 20 vs 30 months, P = 0.37). By comparing multi-omics differences between patient groups, we observed that MMR-related genes were specifically downregulated in IDHwt grade 2 gliomas at both the single-cell and bulk transcriptomic levels. DiscussionThe therapeutic benefit of TMZ in IDHwt histologic grade 2 gliomas appears to be limited, with even a potential adverse impact on survival. Therefore, postoperative use of TMZ as a recommended chemotherapy should be approached with caution in these patients, particularly in cases with unmethylated MGMT-p, where alternative treatment strategies are warranted. High MGMT expression and specific downregulation of MMR-related genes may represent key factors underlying the limited efficacy of TMZ in IDHwt, MGMT-p unmethylation grade 2 gliomas. Level of evidenceIII.

Reliable non-invasive biomarkers for tumor grading and disease monitoring in pediatric brain tumors are an unmet clinical need. Circulating extracellular vesicles (EVs) carrying molecular cargo reflective of tumor biology, offer promise as liquid biopsy tools. We have previously discovered EV-associated Syndecan-1 (SDC1) to be overexpressed in malignant brain tumors, but its value as a biomarker in pediatric disease remains unclear. In this study, plasma EVs were isolated from pediatric brain tumor patients (n=60) by size-exclusion chromatography and characterized using cryo-electron microscopy, nanoflow cytometry, immunoblotting, and single-vesicle total internal reflection fluorescence imaging. EV-associated SDC1 (EV-SDC1) was quantified and analyzed in relation to tumor grade, subtype, surgical resection status, and tumor volume. EV-SDC1 levels were significantly elevated in high-grade (ependymoma, diffuse midline glioma, and atypical teratoid/rhabdoid tumor (AT/RT)) compared with low-grade pilocytic astrocytoma tumors and robustly discriminated grade 3 tumors from pilocytic astrocytoma (AUROC 1.00). Independent validation using transcriptomic data from the Open Pediatric Brain Tumor Atlas showed SDC1 mRNA levels to effectively distinguish high grade (ependymoma, medulloblastoma, diffuse midline glioma, and AT/RT) from pilocytic astrocytoma patients. Furthermore, EV-SDC1 levels decreased following complete tumor resection but remained elevated in patients with residual disease or recurrence. Collectively, circulating SDC1-positive EVs represents a clinically informative biomarker reflecting tumor aggressiveness and treatment response in pediatric brain tumors, supporting their potential for non-invasive disease stratification and monitoring.

BackgroundGlioblastoma (GB) is the most common and most aggressive malignant primary brain tumor, with 5-year survival rates around 5%. While treatment-related factors significantly impact outcomes, the influence of sociodemographic variables remains unclear, with prior studies showing inconsistent findings. These inconsistencies may stem from underrepresentation of rural patients within national databases and from comparison across centers with variable treatment practices. ObjectiveTo evaluate whether population size, geographic distance, and regional-average income influence short- and long-term survival among GB patients when treatment delivery is uniform. MethodsWe analyzed 248 patients who underwent surgical resection at a single publicly funded tertiary neurosurgical center serving all of Southeastern Ontario (2017-2023). Multivariable logistic regression assessed 90-day mortality, while Cox proportional hazards models with time-varying treatment covariates evaluated overall survival (OS). ResultsAmong 248 patients, 71.8% resided in communities with <50,000 residents. Residing in areas with [&ge;]50,000 residents was associated with 63% lower odds of 90-day mortality (OR 0.37; 95% CI: 0.17-0.81). Each 10-mile increase in distance was associated with 5% increased odds of 90-day mortality (OR 1.05; 95% CI: 1.01-1.10). Regional-average income showed no association with 90-day mortality (OR 0.96; 95% CI: 0.51-1.81). None of these variables significantly affected OS in multivariable models. Conclusions: When treatment practices are uniform, population size and geographic distance independently influence early post-operative mortality but not long-term survival. These findings suggest that improving equity in GB care requires targeted interventions during the critical first 90 days post-surgery, extending beyond geographic access alone to address challenges inherent to low-population-density communities. Key PointsO_LIRural residence and greater distance from tertiary care independently increase 90-day mortality after glioblastoma surgery, even when treatment practices are uniform, suggesting distinct mechanisms requiring separate interventions. C_LIO_LIThese sociodemographic factors do not affect overall survival when treatment access is equalized, indicating vulnerability occurs specifically during the early post-operative period rather than throughout the disease course. C_LIO_LIRegional-average income did not influence survival within our universal healthcare system, demonstrating that socioeconomic disparities in GB outcomes may be modifiable through system-level policies ensuring equitable treatment access. C_LI Importance of StudyPersistent outcome disparities between rural and urban glioblastoma patients have long been attributed to unequal access to oncologic therapy. By studying a cohort where all patients received care at the same tertiary center with uniform treatment practices, this study isolates the independent effects of geography from treatment variation. Our findings reveal that population size and distance primarily affect 90-day post-operative mortality rather than long-term survival, fundamentally reframing the mechanism through which geography influences these outcomes. This challenges the assumption that simply bringing care closer to rural communities will eliminate disparities. Instead, health systems must develop comprehensive interventions targeting the critical early post-operative period, including enhanced care coordination, proactive follow-up systems, and community-based support programs tailored to rural populations. The null effect of income in our publicly funded system provides compelling evidence that socioeconomic disparities observed in private healthcare systems are modifiable through universal coverage policies. This methodological approach, combining time-varying treatment analysis, uniform care delivery, and representative sampling of underserved populations, offers a replicable framework for investigating disparities in other cancer types and healthcare contexts.

BackgroundMeningioma-induced hyperostosis (MIH) is a frequent radiographic finding, yet its underlying mechanisms remain poorly understoodWhile hyperostosis has traditionally been treated as a binary phenomenon, the aim of this study was to determine whether MIH represents a heterogenous process with distinct radiological subtypes associated with genetic associations. MethodsWe retrospectively reviewed the records and imaging of patients with meningiomas resected between 2021-2024 at a single institution. Somatic mutations identified through next-generation sequencing were analyzed. CT images were analyzed for bone involvement and hyperostosis subtype. Type I hyperostosis was defined by destruction of cortical architecture while Type II hyperostosis was defined by the preservation of cortical structure. Associations with TRAF7 mutations were assessed using univariate testing, multivariable logistic regression, and supervised machine-learning models. Quantitative bone density analysis was performed using region-of-interest grayscale histogram analysis. ResultsAmong 384 tumors, 54 (14.1%) exhibited hyperostosis--23 Type I and 31 Type II. TRAF7 mutations were significantly enriched in Type I hyperostosis compared with Type II and non-hyperostotic tumors (78.3% vs 25.8% vs 17.0%, p<0.001). Type I hyperostosis independently predicted TRAF7 mutations (OR:18.73, p=0.001), along with skull base location, smaller tumor size, homogeneous contrast enhancement, and extensive T2 hyperintensity. Gradient boosting achieved the highest predictive accuracy (AUC=0.854). Quantitative bone density analysis demonstrated preserved cortical-cancellous architecture in Type II hyperostosis, whereas Type I showed architectural disruption. ConclusionsMIH is a radiographically heterogenous phenomenon. Hyperostosis with disrupted cortical architecture is strongly associated with TRAF7 mutations and may represent a key feature of this mutations radiographic phenotype.

BackgroundMalignant pediatric brain tumors remain the leading cause of cancer-related mortality in children. Current diagnostics rely on imaging and invasive biopsy, which may not capture tumor heterogeneity. Liquid biopsy-based biomarkers such as tumor-educated platelets have shown diagnostic value in adult cancers, but their utility in pediatric brain tumors has not been investigated. MethodsWe analyzed platelet transcriptomes of 73 blood samples from 23 pediatric brain tumor patients, classified as high-grade tumor patients or low-grade tumor patients, and 25 cancer-free controls. Platelets were isolated, CD45 depleted, and RNA was extracted for RNA sequencing. CD45 depletion efficiency was assessed using xCell-based leukocyte enrichment scores. Differential gene expression was assessed with DESeq2 and Gene Ontology over-representation analysis. Gene-level discrimination between groups was evaluated by receiver operating characteristic analysis, and a logistic regression model with patient-grouped 5-fold cross-validation was trained to classify high-grade tumor patients vs. controls. ResultsPlatelets from brain tumor patients showed transcriptional remodeling compared to controls, especially pronounced in high-grade tumor patients. We identified 398 and 649 differentially expressed genes in the brain tumor group vs. controls and high-grade tumor patients vs. control comparisons, respectively, and 85 genes in high-grade tumor patients vs. low-grade tumor patients. No genes met significance in low-grade tumor patients vs. controls. High-grade tumor patients showed consistent upregulation of cancer-associated mitochondrial genes. Similarly, gene enrichment analyses highlighted pathways related to mitotic regulation, chromosome segregation, and mitochondrial metabolism. Multiple genes demonstrated strong diagnostic performance, and logistic regression classifier based on selected platelet transcripts achieved an area under the curve of 0.89, sensitivity of 79%, and a specificity of 80% in identifying high-grade tumor patients. ConclusionsThis study provides the first evidence that platelets exhibit distinct transcriptomic signatures in pediatric brain tumor patients. Platelet RNA profiles robustly differentiate high-grade tumor patients from low-grade tumor patients and cancer-free controls, reflecting tumor presence and biological aggressiveness. These findings support the feasibility of tumor-educated platelets as a minimally invasive biomarker for pediatric malignant brain tumor detection, and longitudinal monitoring. Larger multicenter studies are warranted to validate applicability.

Abstract Background: Conventional MRI cannot reliably distinguish radiation necrosis (RN) from recurrent metastasis after cranial radiotherapy, as both can show similar enhancement despite different biology. We tested whether these entities are mechanically non-equivalent in vivo and separable by MRE-derived viscoelastic metrics and perilesional interface-instability features. Methods: In a prospective, histopathology-anchored cohort, 11 post-radiotherapy enhancing lesions were classified as RN (n=3) or recurrent/progressive tumor (n=8). MRE was acquired at 3.0 T with single-frequency 60-Hz excitation to derive storage modulus (G'), loss modulus (G''), and complex shear modulus magnitude (|G*|). Co-primary endpoints were median tumor G' and |G*|, each tested one-sided (RN > tumor) with Holm correction across the two co-primary tests. Median tumor G'' was tested two-sided. A prespecified secondary 6-endpoint family (absolute and tumor/NAWM-normalized G', G'', and |G*|) was analyzed with Benjamini-Hochberg FDR control. Exploratory instability mapping in a 0- 6 mm peritumoral shell generated interface-topology metrics, including convexity. Results: Absolute tumor-core medians were higher in RN than tumor for |G*| (1.79 vs 1.32 kPa; Cliff's {delta} = 0.67; q = 0.10), G' (1.62 vs 1.09 kPa; {delta} = 0.50; q = 0.14), and G'' (0.81 vs 0.46 kPa; {delta} = 0.75; q = 0.10). NAWM normalization improved separation: tumor/NAWM |G*| (2.26 vs 1.41; {delta} = 0.92; q = 0.04) and tumor/NAWM G'' (2.67 vs 0.87; {delta} = 1.00; q = 0.04) were FDR-significant. Convexity also differentiated RN from tumor (0.49 vs 0.36; {delta} = 1.00; MWU p = 0.01). Conclusions: Tumor/NAWM G'', tumor/NAWM |G*|, convexity, and tumor G'' emerged as the strongest candidate features, indicating that RN is mechanically harder and more dissipative than recurrent metastasis. Signal strength was high (Cliff's {delta} up to 1.00) but should be interpreted cautiously given sample size. Exploratory analyses further suggest that instability mapping captures biologically relevant interface behavior. These findings support a mechanics-based RN-versus-recurrence framework and justify prespecified, preregistered external validation.

BackgroundGlioblastoma (GBM) represents an aggressive malignancy with limited therapeutic options. The immunosuppressive nature of GBM may be reversible with immune checkpoint inhibitor (ICI) treatment, however, initial studies have yet to demonstrate this. It is postulated that trafficking of peripherally activated lymphocytes may play a role in generating a robust intracranial immune response. Therefore, a blood-based assay to identify peripheral blood response may both predict response and better identify the ideal patient populations for future ICI clinical trials. MethodsThis was an open-label, Phase II, investigator-initiated exploratory study of patients with newly diagnosed GBM who completed maximal tumor resection and concurrent chemoradiation followed by standard adjuvant temozolomide and the combination of Nivolumab and Ipilimumab. The primary objective was to determine if the outcome, as measured by overall survival, is improved in patients when treatment with immune checkpoint inhibitors results in an immune response in peripheral blood T lymphocytes. The immune response is defined as changes in the CD4+/CD8+ precursor frequency and expansion index compared to the overall survival (OS) measured in months. ResultsThe study closed to enrollment early due to a shift in clinical priorities after the accrual of 40 patients. Twenty-three patients have died of their disease, and adequate samples for the primary analysis were available for 17 of these patients. The median OS for the 17 patients was 19 months (range 9-45months). For the four immune measurements, patients were categorized as reactive, indeterminate, or suppressed based on pre-defined protocol criteria. Only two patients were classified as reactive across all four measurements, and their median OS was 17.5 months, compared with 21 months for patients classified as suppressed. Across each of the 4 individual immune measurements, no statistical difference in OS were observed between reactive and suppressed groups. ConclusionIn this limited cohort, no detectable difference in the OS was observed between patients with a reactive immune signature and those with a suppressed immune signature.

BackgroundPrimary central nervous system (CNS) tumors affect patients psychological well-being and quality of life. Individualized approaches, such as Managing Cancer and Living Meaningfully (CALM), have shown potential in advanced cancers for improving these outcomes. AimsThis study assessed the effects and feasibility of CALM delivered remotely to a diverse cohort of patients with a primary CNS tumor. MethodsPatients completed 3-6 remote CALM sessions focusing on 4 interrelated domains. Depression, death anxiety, attachment style, and quality of life were assessed at study enrollment, 3-months, and 6-months into the intervention. ResultsOf the 19 patients enrolled, 15 (79% retention rate) completed the study. Most patients had a high-grade (47%) tumor, mainly diagnosed in the brain (60%). The median age was 44 years (range, 24-70). Feasibility was demonstrated through adherence to completing outcome questionnaires and a high level of patient satisfaction (100% found it worthwhile). Although no statistically significant changes were seen in depression, death anxiety, attachment anxiety, or quality of life (p > 0.05; g = -0.09 to 0.78) at any measured time, a clinically meaningful decrease in depression was observed at the 6-month point (mean difference = -3.36, p = 0.13) among spine tumor patients. ConclusionsThis study demonstrated that delivering CALM via telehealth is feasible, as evidenced by high compliance, low attrition, and acceptability among patients diagnosed with CNS tumors. The findings indicated meaningful reductions in depressive symptoms among patients with spinal cord tumors. These preliminary positive findings justify further evaluation of the feasibility and effectiveness of CALM in a larger sample. Trial registrationClinicalTrials.gov ID NCT04852302

BackgroundPrimary malignant brain tumors (PMBTs), including glioblastomas (GBM) and low-grade gliomas (LGGs), frequently cause seizures, which worsen patient morbidity and quality of life. Alcohol use disorder (AUD) is a known risk factor for seizures in the general population, but its role in PMBT-associated seizures remains poorly understood. ObjectiveTo evaluate the association between AUD and seizure prevalence in PMBT patients, and to examine whether this relationship varies by age, sex, race/ethnicity, and tumor grade. MethodsWe conducted a retrospective cohort study using the Cosmos EHR database (2015-2025), identifying 244,955 adult patients with PMBTs. Seizure/epilepsy and AUD diagnoses were ascertained via ICD-10 codes. Associations were assessed using chi-square tests and logistic regression (adjusting for age, sex, and race). Temporal sequencing of AUD and seizure diagnoses was analyzed in six-month intervals. Subgroup analyses were performed for glioma grade (low vs. high). ResultsPMBT patients with seizures were nearly twice as likely to have AUD compared to those without (OR = 1.90, 95% CI 1.83-1.97, p < 0.001). After adjusting for demographics, AUD remained significantly associated with seizure risk (OR = 1.37, 95% CI 1.24-1.51, p < 0.001). This association was strongest in younger patients and present across all sexes and racial groups. Temporal analyses indicated that AUD partly preceded seizure onset. Among patients with available histologic data, alcohol-using low-grade patients exhibited a markedly higher seizure prevalence compared to high-grade patients (43.8% vs. 12.1%, p < 0.001). ConclusionsAUD is potentially associated with increased seizure risk in PMBT patients, with a stronger effect in younger individuals and low-grade tumors. These findings suggest that alcohol-related hyperexcitability may compound tumor-associated epileptogenesis, highlighting AUD as a potentially modifiable risk factor. Prospective studies are warranted to confirm these relationships and to evaluate interventions targeting alcohol use for seizure mitigation. Key PointsO_LIPMBT patients with seizures had higher AUD prevalence (OR = 1.90, 95% CI 1.83-1.97). C_LIO_LIModels controlling for demographic variables confirmed increased seizure risk persisted in the AUD population (OR = 1.37, 95% CI 1.24-1.51). C_LIO_LIAssociation strongest in younger patients, unchanged across sex and race groups. C_LIO_LISeizures were more common in AUD low-grade gliomas vs high-grade (43.8% vs 12.1%). C_LIO_LITemporal data suggest AUD partly preceded seizure onset in PMBT patients. C_LI

Brain tumors are one of the most life-threatening diseases, requiring precise and timely detection for effective treatment. Traditional methods for brain tumor detection rely heavily on manual analysis of MRI scans, which is time-consuming, subjective, and prone to human error. With advancements in deep learning, Convolutional Neural Networks (CNNs) have become popular for medical image analysis. However, CNNs are limited in their ability to capture spatial hierarchies and pose variations, which reduces their accuracy, particularly for tasks like brain tumor segmentation where precise spatial relationships are crucial. This research introduces a hybrid Capsule Neural Network (CapsNet) and ResNet50 model designed to overcome the limitations of traditional CNNs by capturing both spatial and pose information in MRI scans. The proposed model leverages ResNet50 for feature extraction and CapsNet for handling spatial relationships, leading to more accurate segmentation. The study evaluates the model on the BraTS2020 dataset and compares its performance to state-of-the-art CNN architectures, including U-Net and pure CNN models. The hybrid model, featuring a custom 5-cycle dynamic routing algorithm to enhance capsule agreement for tumor boundaries, achieved 98% accuracy and an F1-score of 0.87, demonstrating superior performance in detecting and segmenting brain tumors. This study pioneers the systematic evaluation of the ResNet50 + CapsNet hybrid on the BraTS2020 dataset, with a tailored class weighting scheme addressing class imbalance, improving effectiveness in identifying irregularly shaped tumors and smaller regions in identifying irregularly shaped tumors and smaller tumor regions. The study offers a robust solution for automating brain tumor detection. Future work will explore the use of Capsule Networks alone for brain tumor detection in MRI data and investigate alternative Capsule Network architectures, as well as their integration into clinical decision support systems.

Lower-grade gliomas (World Health Organization [WHO] grades 2-3) exhibit variable treatment responses, yet clinical decisions remain guided by population-level trial results. Standard causal survival forests estimate treatment effects at individual time horizons but lack methodology to synthesize these into interpretable temporal trajectories. Here, we apply the Causal Analysis of Survival Trajectories (CAST) framework, a recently developed extension of causal survival forests that synthesizes horizon-specific causal effect estimates into smooth temporal curves while accounting for between-horizon covariances via bootstrap estimation and Ledoit-Wolf shrinkage. We apply CAST to estimate time-varying, heterogeneous effects of radiotherapy and chemotherapy in 776 patients with lower-grade gliomas from The Cancer Genome Atlas (TCGA; n=512) and the Chinese Glioma Genome Atlas (CGGA; n=264), analyzing six treatment-outcome scenarios and adjusting for age, sex, WHO grade, isocitrate dehydrogenase (IDH) mutation status, 1p/19q codeletion, and extent of resection using elastic net propensity scores with overlap weighting. CAST curves reveal that chemotherapy provides consistent, sustained benefits across both cohorts; survival probability gains peak at 0.31 at 72-84 months for TCGA overall survival and 0.46 at 48 months for progression-free survival, with restricted mean survival time gains of 18.4 and 32.5 months at 10 years, respectively. CGGA chemotherapy shows delayed but large positive effects (survival probability peak 0.48 at 108 months). Radiotherapy effects are mixed, with modest E-values indicating sensitivity to residual confounding by indication. Subgroup CAST curves identify age at diagnosis as the dominant driver of treatment effect heterogeneity (46-56% of splits). All findings are robust to placebo permutation, simulated unobserved confounder, and negative control refutation tests. The CAST framework provides a general-purpose tool for temporal treatment effect visualization applicable beyond neuro-oncology.

Glioblastoma is increasingly treated with Tumor Treating Fields (TTFields), but how post-craniotomy anatomy and fixation hardware alter delivered fields is unclear. We used finite-element modeling in a realistic head model to simulate TTFields after a standard bone flap with either a non-penetrating fixation plate or a penetrating skull clamp, and compared results to an intact-skull baseline across a range of clinically used array layouts. Bone gaps increased mean brain electric-field magnitude by [~]10-20%. Non-penetrating plates caused only minimal, localized changes relative to the bone-gap condition. In contrast, penetrating clamps produced strong but spatially confined increases: local mean fields were [~]6-8x higher within 5-10 mm of the device, with [&ge;] 50% enhancement extending [~]50-60 mm depending on whether the gap was modeled as healed scalp (soft-tissue-like) or healed bone; this enhancement decayed with distance. These simulations, performed in a single head model with literature-based tissue conductivities, suggest that penetrating hardware can substantially modulate local TTFields delivery, whereas non-penetrating plates have minimal impact. Accounting for post-surgical anatomy and hardware in TTFields planning may improve dose targeting.

BackgroundAccurate contouring of target volumes and organs at risk is critical for radiotherapy. While deep learning (DL) models offer efficient automation, their generalizability to real-world clinical cases containing anatomical variations and artifacts requires rigorous validation. PurposeTo evaluate the clinical accuracy and robustness of RatoGuide, a novel DL-based auto-segmentation software, using a dataset derived from routine clinical practice including atypical cases. MethodsThis single-center retrospective study included 36 patients treated for head and neck, thoracic, abdominal, and pelvic cancers. The cohort was intentionally selected to encompass diverse anatomies and artifacts (e.g., pacemakers, artificial femoral head replacement). Auto-contours generated by RatoGuide were compared with expert-approved manual contours. Performance was evaluated quantitatively using the Dice Similarity Coefficient (DSC) and 95th percentile Hausdorff Distance (HD95), and qualitatively via a 5-point visual assessment scale (higher is better) by four independent reviewers. A score of [&le;]2 by multiple reviewers was defined as failure. ResultsOverall, the mean DSC, HD95, and visual assessment score were 0.79 {+/-} 0.19, 6.35 {+/-} 12.2 mm, and 3.65 {+/-} 0.88, respectively. The mean DSC exceeded 0.8 in 62% (23/37 organ structures) of the evaluated structure types, and a total of 93.5% (315/337) of all contours were considered clinically acceptable based on visual evaluation . However, lower performance was observed in small structures (e.g., optic chiasm) and low-contrast organs (e.g., esophagus). ConclusionsRatoGuide demonstrated favorable performance for major organs across various anatomical regions, consistent with benchmarks reported in the literature. However, performance variability in atypical cases underscores the necessity of rigorous visual verification by experts for clinical implementation.

IntroductionIdiopathic normal pressure hydrocephalus (iNPH) is a partially reversible neurological disorder in which imaging biomarkers support diagnosis and surgical decision-making. The callosal angle (CA) is one of the most robust radiological markers of iNPH and has also been associated with postoperative shunt outcome. However, several manual measurement variants exist and artificial intelligence (AI)-based tools now enable automatic CA measurement. Materials and MethodsIn total 71 patients (40 with confirmed iNPH and 31 controls) were included. Six predefined manual methods for measuring CA were applied to preoperative 3D T1-weighted MRI and evaluated for diagnostic performance and interobserver agreement. An AI-derived automatic CA (cMRI from Combinostics) was included as a seventh method and compared with the traditional manual method (perpendicular to the bicommissural plane and through the posterior commissure). Automatic measurements were additionally assessed in pre- and postoperative scans to evaluate robustness against shunt-related artifacts. ResultsAll seven CA variants significantly differentiated iNPH patients from controls (p < 0.05). The traditional method showed the highest discriminative performance (AUC = 0.986, SE = 0.012), while alternative planes demonstrated slightly lower accuracy (AUC range = 0.957-0.978). Interobserver agreement for manual measurements was good to excellent (ICC = 0.687-0.977). Automatic CA measurements showed excellent correlation with the traditional method, preoperative ICC = 0.92; postoperative ICC = 0.96. ConclusionAlthough several CA positions perform comparably, the traditional method remains marginally superior and is best supported by the literature. Automated CA measurements closely match expert manual assessment in pre- and postoperative imaging, supporting clinical implementation.