Explore the vast range of titles available.

- In the recently updated World Health Organization (WHO) classification of central nervous system tumors, our concept of infiltrating gliomas as a molecular dichotomy between oligodendroglial and astrocytic tumors has been codified. Advances in animal models of glioma and a wealth of sophisticated molecular analyses of human glioma tissue have led to a greater understanding of some of the biologic underpinnings of gliomagenesis. - To review our understanding of gliomagenesis in the setting of the recently updated WHO classification of central nervous system tumors. Topics addressed include a summary of an updated diagnostic schema for infiltrating gliomas, the crucial importance of isocitrate dehydrogenase mutations, candidate cells of origin for gliomas, environmental and other posited contributing factors to gliomagenesis, and the possible role of chromatin topology in setting the stage for gliomagenesis. - We conducted a primary literature search using PubMed. - With multidimensional molecular data sets spanning increasingly larger numbers of patients with infiltrating gliomas, our understanding of the disease at the point of surgical resection has improved dramatically and this understanding is reflected in the updated WHO classification. Animal models have demonstrated a diversity of candidates for glioma cells of origin, but crucial questions remain, including the role of neural stem cells, more differentiated progenitor cells, and glioma stem cells. At this stage the increase in data generated from human samples will hopefully inform the creation of newer animal models that will recapitulate more accurately the diversity of gliomas and provide novel insights into the biologic mechanisms underlying tumor initiation and progression.

While Machine Learning (ML) models have been increasingly applied to a range of histopathology tasks, there has been little emphasis on characterizing these models and contrasting them with human experts. We present a detailed empirical analysis comparing expert neuropathologists and ML models at predicting IDH mutation status in H&E-stained histology slides of infiltrating gliomas, both independently and synergistically. We find that errors made by neuropathologists and ML models trained using the TCGA dataset are distinct, representing modest agreement between predictions (human-vs.-human κ = 0.656; human-vs.-ML model κ = 0.598). While no ML model surpassed human performance on an independent institutional test dataset (human AUC = 0.901, max ML AUC = 0.881), a hybrid model aggregating human and ML predictions demonstrates predictive performance comparable to the consensus of two expert neuropathologists (hybrid classifier AUC = 0.921 vs. two-neuropathologist consensus AUC = 0.920). We also show that models trained at different levels of magnification exhibit different types of errors, supporting the value of aggregation across spatial scales in the ML approach. Finally, we present a detailed interpretation of our multi-scale ML ensemble model which reveals that predictions are driven by human-identifiable features at the patch-level.

Anna Lasorella, Raul Rabadan, Antonio Iavarone and colleagues report an integrated analysis of genomic alterations in glioblastoma. They identify and functionally validate several new driver events, including loss-of-function mutations in CTNND2 and recurrent EGFR fusions. Glioblastoma is one of the most challenging forms of cancer to treat. Here we describe a computational platform that integrates the analysis of copy number variations and somatic mutations and unravels the landscape of in-frame gene fusions in glioblastoma. We found mutations with loss of heterozygosity in LZTR1 , encoding an adaptor of CUL3-containing E3 ligase complexes. Mutations and deletions disrupt LZTR1 function, which restrains the self renewal and growth of glioma spheres that retain stem cell features. Loss-of-function mutations in CTNND2 target a neural-specific gene and are associated with the transformation of glioma cells along the very aggressive mesenchymal phenotype. We also report recurrent translocations that fuse the coding sequence of EGFR to several partners, with EGFR-SEPT14 being the most frequent functional gene fusion in human glioblastoma. EGFR-SEPT14 fusions activate STAT3 signaling and confer mitogen independence and sensitivity to EGFR inhibition. These results provide insights into the pathogenesis of glioblastoma and highlight new targets for therapeutic intervention.

Glioblastomas (GBMs) diffusely infiltrate the brain, making complete removal by surgical resection impossible. The mixture of neoplastic and nonneoplastic cells that remain after surgery form the biological context for adjuvant therapeutic intervention and recurrence. We performed RNA-sequencing (RNA-seq) and histological analysis on radiographically guided biopsies taken from different regions of GBM and showed that the tissue contained within the contrast-enhancing (CE) core of tumors have different cellular and molecular compositions compared with tissue from the nonenhancing (NE) margins of tumors. Comparisons with the The Cancer Genome Atlas dataset showed that the samples from CE regions resembled the proneural, classical, or mesenchymal subtypes of GBM, whereas the samples from the NE regions predominantly resembled the neural subtype. Computational deconvolution of the RNA-seq data revealed that contributions from nonneoplastic brain cells significantly influence the expression pattern in the NE samples. Gene ontology analysis showed that the cell type-specific expression patterns were functionally distinct and highly enriched in genes associated with the corresponding cell phenotypes. Comparing the RNA-seq data from the GBM samples to that of nonneoplastic brain revealed that the differentially expressed genes are distributed across multiple cell types. Notably, the patterns of cell type-specific alterations varied between the different GBM subtypes: the NE regions of proneural tumors were enriched in oligodendrocyte progenitor genes, whereas the NE regions of mesenchymal GBM were enriched in astrocytic and microglial genes. These subtypespecific patterns provide new insights into molecular and cellular composition of the infiltrative margins of GBM.

Craniosynostosis is a group of disorders of premature calvarial suture fusion. The identity of the calvarial stem cells (CSCs) that produce fusion-driving osteoblasts in craniosynostosis remains poorly understood. Here we show that both physiologic calvarial mineralization and pathologic calvarial fusion in craniosynostosis reflect the interaction of two separate stem cell lineages; a previously identified cathepsin K (CTSK) lineage CSC 1 (CTSK + CSC) and a separate discoidin domain-containing receptor 2 (DDR2) lineage stem cell (DDR2 + CSC) that we identified in this study. Deletion of Twist1 , a gene associated with craniosynostosis in humans 2 , 3 , solely in CTSK + CSCs is sufficient to drive craniosynostosis in mice, but the sites that are destined to fuse exhibit an unexpected depletion of CTSK + CSCs and a corresponding expansion of DDR2 + CSCs, with DDR2 + CSC expansion being a direct maladaptive response to CTSK + CSC depletion. DDR2 + CSCs display full stemness features, and our results establish the presence of two distinct stem cell lineages in the sutures, with both populations contributing to physiologic calvarial mineralization. DDR2 + CSCs mediate a distinct form of endochondral ossification without the typical haematopoietic marrow formation. Implantation of DDR2 + CSCs into suture sites is sufficient to induce fusion, and this phenotype was prevented by co-transplantation of CTSK + CSCs. Finally, the human counterparts of DDR2 + CSCs and CTSK + CSCs display conserved functional properties in xenograft assays. The interaction between these two stem cell populations provides a new biologic interface for the modulation of calvarial mineralization and suture patency. The calvarial stem cell niche is populated by a cathepsin K-expressing cell lineage and a newly identified discoidin domain-containing receptor 2-expressing lineage, both of which are required for proper calvarial mineralization.

Large Language Models (LLMs) have shown significant promise across various natural language processing tasks. However, their application in the field of pathology, particularly for extracting meaningful insights from unstructured medical texts such as pathology reports, remains underexplored and not well quantified. In this project, we leverage state-of-the-art language models, including the GPT family, Mistral models, and the open-source Llama models, to evaluate their performance in comprehensively analyzing pathology reports. Specifically, we assess their performance in cancer type identification, AJCC stage determination, and prognosis assessment, encompassing both information extraction and higher-order reasoning tasks. Based on a detailed analysis of their performance metrics in a zero-shot setting, we developed two instruction-tuned models: Path-llama3.1-8B and Path-GPT-4o-mini-FT . These models demonstrated superior performance in zero-shot cancer type identification, staging, and prognosis assessment compared to the other models evaluated.

Histopathology is a vital diagnostic discipline in medicine, fundamental to our understanding, detection, assessment and treatment of conditions such as cancer, dementia and heart disease. Traditionally, the standard workflow in histopathology has primarily relied on the visual interpretation of tissue samples carried out by human experts under a light microscope. Since the 2000s, thanks to advances in scanning technologies such as whole-slide imaging, histopathology is undergoing a digital transformation. The rapid increase in digital data is fuelling the development and application of artificial intelligence (AI) methods. In this Review, we delve into the latest progress in AI methods for histopathology, which promise to yield accurate, scalable, useful and affordable support tools for clinical decision. We examine the challenges and opportunities in this domain, exploring historically important approaches and problems that have shaped the field, while also highlighting recent technological breakthroughs that are poised to redefine its future. Furthermore, we offer an overview of publicly available datasets that have been instrumental in propelling the development of AI methods in histopathology.Increase in clinical digital data is propelling the development and application of artificial intelligence methods in histopathology. In this Review, machine learning algorithms and models and their clinical use cases are discussed, highlighting the computational and operational challenges in the field.

The molecular characteristics of metastatic upper tract urothelial carcinoma (UTUC) are not well understood, and there is a lack of knowledge regarding the genomic and transcriptomic differences between primary and metastatic UTUC. To address these gaps, we integrate whole-exome sequencing, RNA sequencing, and Imaging Mass Cytometry using lanthanide metal-conjugated antibodies of 44 tumor samples from 28 patients with high-grade primary and metastatic UTUC. We perform a spatially-resolved single-cell analysis of cancer, immune, and stromal cells to understand the evolution of primary to metastatic UTUC. We discover that actionable genomic alterations are frequently discordant between primary and metastatic UTUC tumors in the same patient. In contrast, molecular subtype membership and immune depletion signature are stable across primary and matched metastatic UTUC. Molecular and immune subtypes are consistent between bulk RNA-sequencing and mass cytometry of protein markers from 340,798 single cells. Molecular subtypes at the single-cell level are highly conserved between primary and metastatic UTUC tumors within the same patient. Detailed molecular studies are required to understand the differences between primary and metastatic upper tract urothelial carcinoma (UTUC). Here, the authors use genomics, transcriptomics and imaging mass cytometry to characterise the molecular profiles of primary and metastatic UTUC, and find that molecular subtypes remain highly conserved.

Since the introduction of integrated histological and molecular diagnoses by the 2016 World Health Organization (WHO) Classification of Tumors of the Nervous System, an increasing number of molecular markers have been found to have prognostic significance in infiltrating gliomas, many of which have now become incorporated as diagnostic criteria in the 2021 WHO Classification. This has increased the applicability of targeted-next generation sequencing in the diagnostic work-up of neuropathology specimens and in addition, raises the question of whether targeted sequencing can, in practice, reliably replace older, more traditional diagnostic methods such as immunohistochemistry and fluorescence in-situ hybridization. Here, we demonstrate that the Oncomine Cancer Gene Mutation Panel v2 assay targeted-next generation sequencing panel for solid tumors is not only superior to IHC in detecting mutation in IDH1/2 and TP53 but can also predict 1p/19q co-deletion with high sensitivity and specificity relative to fluorescence in-situ hybridization by looking at average copy number of genes sequenced on 1p, 1q, 19p, and 19q. Along with detecting the same molecular data obtained from older methods, targeted-next generation sequencing with an RNA sequencing component provides additional information regarding the presence of RNA based alterations that have diagnostic significance and possible therapeutic implications. From this work, we advocate for expanded use of targeted-next generation sequencing over more traditional methods for the detection of important molecular alterations as a part of the standard diagnostic work up for CNS neoplasms.