Morana Vojnic

Purpose For over a century, the pathologist’s interpretation at the microscope was the final word on a brain tumor’s identity. The 2021 World Health Organization (WHO) classification of central nervous system (CNS) tumors, 5th edition, changed that hierarchy by introducing a layered, integrated diagnosis in which molecular findings take precedence over histology when the two conflict. Artificial intelligence (AI) applied to digitized slides adds a computational layer to this framework and marks a fundamental change in diagnostic workflow. This review examines the role of AI and digital pathology in adult CNS tumors and discusses the opportunities and challenges in clinical, neuro-oncology implementation. Methods We conducted a narrative review of relevant literature indexed in PubMed before and after the 2021 WHO classification, covering digital pathology, whole-slide imaging (WSI), and AI applications in adult neuro-oncology. We synthesized evidence on diagnostic performance, molecular prediction, prognostic modeling, workflow integration, and ethical considerations. Results AI and digital pathology show potential to improve diagnostic consistency, shorten turnaround time, and integrate histopathological, molecular, and radiological information. AI models may support glioma grading, prediction of molecular alterations from routine histology, survival stratification, prediction of brain metastasis, and intraoperative decisions. Barriers to clinical adoption include limited generalizability, lack of transparency, data privacy, and evolving regulatory oversight. Conclusion AI and digital pathology are transforming adult CNS tumor diagnostics and could improve precision, efficiency, and multidisciplinary decision-making. Successful implementation will require rigorous multi-institutional clinical validation, transparent governance, equitable access, and collaboration among patients and health care professionals.

Purpose: O^6^-methylguanine–DNA methyltransferase (MGMT) promoter methylation is a key predictive biomarker for temozolomide (TMZ) response in glioma. MGMT status is routinely assessed at diagnosis; however, its utility at recurrence remains incompletely characterized. We aimed to quantify MGMT methylation stability and assess the clinical relevance of reported status changes in recurrent glioma. Methods: We analyzed paired tumor samples from 426 patients with recurrent glioma in a large commercial molecular database containing quantitative methylation percentage and binary methylation status at two sample time points. A separate multi-institutional clinical cohort ( n = 27) with detailed treatment annotation was analyzed to explore associations between MGMT status at recurrence and treatment decision-making. Statistical analyses assessed quantitative methylation dynamics and time between samples as a surrogate for disease course. Results: MGMT status remained unchanged in 84% of tumors across recurrence. 16% demonstrated a status change, commonly among tumors with low baseline methylation near classification thresholds. Longitudinal quantitative methylation changes were modest overall. Tumors that remained, gained, or lost hypermethylation demonstrated longer intervals between samples compared to consistently unmethylated tumors ( p < 0.01). In the clinical cohort, TMZ rechallenge was more frequently pursued in patients retaining hypermethylation at recurrence. Conclusion: MGMT promoter methylation is quantitatively stable in the majority of recurrent gliomas. Status changes predominantly occur in tumors with borderline methylation, highlighting the limitations of binary classification. These findings support routine reassessment of MGMT at recurrence and underscore the value of quantitative reporting to better inform therapeutic decision-making, particularly regarding TMZ rechallenge.

Background: Diffuse midline glioma (DMG) with the H3K27M mutation is an aggressive WHO grade 4 central nervous system tumor characterized by diffuse infiltration of midline structures and a poor prognosis. Molecular characterization is essential for diagnosis and treatment planning; however, obtaining tissue can be infeasible, particularly in anatomically inaccessible locations or medically complex situations, like pregnancy. Patient presentation: A 34-year-old G2P1 woman at 19 weeks’ gestation presented with progressive right-sided numbness and weakness. Magnetic resonance imaging demonstrated an infiltrative lesion in the left midbrain extending to the pons and thalamus with intralesional hemorrhage. Differential considerations included diffuse midline glioma and inflammatory demyelinating disease. Molecular findings: Because of the lesion location and concerns regarding stereotactic biopsy during pregnancy, cerebrospinal fluid (CSF) next-generation sequencing was pursued. CSF profiling identified an H3K27M mutation, establishing the diagnosis of DMG. Management and outcome: Following multidisciplinary discussion involving neuro-oncology, neurology, neurosurgery, radiation oncology, and maternal-fetal medicine, the patient underwent fractionated radiotherapy beginning at 22 weeks’ gestation using pregnancy-shielding techniques. Postpartum dordaviprone (ONC201) therapy was initiated and well tolerated. Follow-up imaging demonstrated radiographic improvement and neurological stabilization, with ongoing response at 12 months post-diagnosis. Conclusions: The case highlights the utility of CSF liquid biopsy in establishing a molecular diagnosis of diffuse midline glioma when tissue biopsy is contraindicated or deferred. The case also demonstrates how molecular profiling can guide precision oncology in medically complex situations like pregnancy.

Background: Pleomorphic xanthoastrocytomas (PXAs) are rare primary central nervous system (CNS) tumors that appear heterogeneous on imaging and histology and typically cause headaches or seizures on initial presentation. Alongside high rates of favorable prognosis after surgical excision exist similarly high rates of recurrence. Leptomeningeal spread on recurrence is even rarer and more challenging to diagnose. Case: We describe a case of a 40‐year‐old man with a history of surgically resected PXA presenting 12 years later with persistent headaches and lower back pain. Imaging studies revealed arachnoiditis, and a subsequent brain biopsy was nondiagnostic. Serial CSF studies only revealed the presence of atypical cells too few to further characterize via standard histology studies, with rare small lymphocytes and monocytoid cells. Submitting these cells for next‐generation sequencing ultimately revealed a BRAF V600E mutation typically found in PXAs, thereby confirming the diagnosis of leptomeningeal recurrence and revealing a therapeutic target. Conclusion: This case highlights the utility of next‐generation sequencing as a means of non‐invasively diagnosing leptomeningeal disease in recurrent PXA and potentially in other cancer types as well.

Background: Leptomeningeal disease (LMD) is characterized by the spread of cancer to the leptomeninges and cerebrospinal fluid (CSF) and is associated with poor survival. Diagnosing LMD is challenging, as traditional methods such as MRI and CSF cell cytology demonstrate variable sensitivity. This study aims to explore the diagnostic potential of CSF circulating tumor cell (CTC) quantification for LMD in conjunction with current standards. Methods: This retrospective case series includes 12 patients with suspected LMD who underwent CSF analysis with the CNSide assay for CTC quantification and next-generation sequencing (NGS), alongside MRI and CSF cytology. Relying on a composite definition of LMD-positive, the diagnostic performance of CTC quantification was assessed. Results: Of 12 patients evaluated for LMD with CNSide, 11 were found to have brain metastases (BM). Lung carcinoma was the most common primary cancer (4/12). Nine patients were deemed LMD-positive based on clinical criteria: 7/9 had a preceding brain metastasis diagnosis, while 2/9 co-presented with LMD and BM. CNSide detected CTCs in 7/12 patients (7/9 of those with LMD) and influenced clinical decision-making by guiding chemotherapy selection and prompting proton craniospinal irradiation. Of the three patients deemed LMD-negative with clinical criteria, all three had negative results on the CNSide assay. The CNSide assay demonstrated a sensitivity of 77.8%, specificity of 100%, and overall accuracy of 83.3%. Conclusion: The integration of CTC quantification with next-generation sequencing (NGS) can be a valuable adjunct to cell cytology in diagnosing leptomeningeal disease. CSF liquid biopsy may provide earlier detection and inform treatment decisions, ultimately improving patient outcomes.

Glioblastoma is an aggressive and treatment-refractory primary brain tumor with limited therapeutic options and high recurrence. The molecular heterogeneity of glioblastoma poses a significant challenge to therapeutic development, as targeted therapies have mostly failed in small-scale clinical trials, underscoring the need for comprehensive next-generation sequencing (NGS) characterization to identify mechanisms of resistance. In this study, we identify and functionally characterize a novel amplified fusion, MDM2 (exon 1)::PDGFRA (exon 8), mediating resistance to cetuximab in an EGFR-amplified glioblastoma. The fusion results in a truncated PDGFRA isoform, in vitro assays demonstrate that MDM2::PDGFRA acts as a constitutively active oncogenic driver with a distinct sensitivity profile to tyrosine kinase inhibitors. Analysis of a glioblastoma cohort indicates PDGFRA structural variants often co-occur with amplification and may serve as biomarkers. These findings highlight the importance of repeat NGS profiling in clinical management and provide a translational framework for identifying and targeting emergent fusion-driven alterations.

Table S1. Complete clinical, pathologic, and molecular annotations of oncogenic MET fusions Table S2. Summary of reports describing the use of MET-TKI to treat MET fusion-positive malignancies Table S3. Summary of antibodies used for Western blot and immunohistochemistry Table S4. List of genes targeted by DNA and RNA sequencing of 56 tumors with MET fusions Table S5. List of primers and plasmid systems used for transfection experiments Table S6. List of mutations detected in MET fusion-positive tumors Table S7. List of copy number alterations detected in MET fusion-positive tumors Table S8. Additional intergenic structural variants, other than MET, detected in MET fusion-positive tumors

Supplementary Figure S1. Features of intergenic MET structural variants (SV) classified as SV of unknown significance Supplementary Figure S2. Clinicopathologic features and mutation signature analysis Supplementary Figure S3. Analysis of MET structural variants and fusions in the GENIE cohort v15.1 Supplementary Figure S4. Experimental modeling of ST7::MET revealed retained ligand dependent activity Supplementary Figure S5. Single-cell expression of detected MET fusion partner genes in normal cells, highlighting resident cells of primary sites of MET-fusion-driven tumors Supplementary Figure S6. Clinical vignette: cholangiocarcinoma with MET, HGF, and EGFR co-amplification and nonamplified ST7::MET detected by RNA only (DNA−/RNA+) Supplementary Figure S7. Extended Main Figure 5 Supplementary Figure S8. Summary of MET fusion-positive tumors treated with MET-TKI reported in the literature Supplementary Figure S9. Transfection of CDH1::MET induced TKI resistance in MET exon 14 skipping mutation-positive Hs746T cells Supplementary Figure S10. Confidence models of ZKSCAN1::MET and SYNGR2::MET homodimers depicted in Figures S7A and 6A, respectively

MET fusions (MET-F) are oncogenic drivers that remain poorly characterized. Analysis of 56 MET-F–positive tumors from an institutional cohort of 91,119 patients (79,864 DNA sequencing plus 11,255 RNA sequencing) uncovered two forms of MET-F pathobiology. The first group featured 5′ partners with homodimerization domains fused in-frame with the MET tyrosine kinase domain, primarily originated from translocations, frequently excluded MET exon 14, mediated oncogenesis through cytoplasmic aggregation and constitutive activation, and were markedly sensitive to MET tyrosine kinase inhibitors (TKI) in preclinical models and patients with lung cancer. The second group lacked partner homodimerization motifs and retained MET transmembrane and extracellular domains. Their pathogenesis involved intrachromosomal rearrangements, resulting in partner selection for promoter hijacking and fusion allele amplification. Membrane-bound fusions were enriched in gliomas with receptor tyrosine kinase co-alterations. We provide a framework to comprehend the heterogeneous landscape of MET-Fs, supporting that fusion oncogenicity and MET TKI sensitivity are determined by structural topology and pathogenomic context. Significance: MET fusions are primary drivers of tumor growth in multiple tumor types – lung cancer and gliomas – and can be effectively targeted with either type I (crizotinib, capmatinib, tepotinib, and savolitinib) or type II (cabozantinib) MET TKIs, with best responses in tumors harboring fusions with partner homodimerization.

This systematic review aims to synthesize the current literature on the association between chemotherapy (CTX) and chemotherapy-related cognitive impairment (CRCI) with functional and structural brain alterations in patients with noncentral nervous system cancers. A comprehensive search of the PubMed/MEDLINE, Web of Science, and Embase databases was conducted, and results were reported following preferred reporting items for systematic review and meta-analyses guidelines. Data on study design, comparison cohort characteristics, patient demographics, cancer type, CTX agents, neuroimaging methods, structural and functional connectivity (FC) changes, and cognitive/psychological assessments in adult patients were extracted and reported. Study quality was assessed using an adapted version of the Newcastle-Ottawa Scale (NOS) for observational studies. Qualitative synthesis of cognitive and psychological testing outcomes, functional and structural connectivity changes, and their associations with CRCI were performed. From 11,335 records identified, 63 studies analyzing 3,642 patients were included. Study designs included 24 prospective studies, 1 retrospective study, 36 cross-sectional studies, and 2 longitudinal studies. Most studies (75%) focused on patients with breast cancer. Common neuroimaging techniques included functional magnetic resonance imaging and diffusion tensor imaging. Postchemotherapy, many studies reported structural and FC alterations in brain networks such as the default mode, central executive, and dorsal attention networks. Cognitive function was assessed in 56 of the 63 included studies. Of the studies examining specific cognitive domains, 64% reported worsened learning and memory, 56% found impaired processing speed, and 70% identified deficits in attention/working memory in patients after CTX. Of the studies examining associations between connectivity changes and worsened cognitive function, 72% reported significant correlations in postchemotherapy patients. However, most of these studies were of low evidence, while 45% of high evidence-level studies, including prospective cohort studies, did not find significant associations between connectivity alterations and cognitive impairments. While there is evidence suggesting CTX affects brain connectivity and neural network dynamics that can lead to cognitive difficulties, the findings are inconsistent. More robust and standardized research is needed to conclusively determine the extent of these effects and to develop targeted interventions for mitigating potential cognitive impairments in patients undergoing systemic treatment.