Glioblastomas (GBMs), among the most aggressive and resilient brain tumors, characteristically exhibit high angiogenic potential, leading to the formation of a dense yet aberrant vasculature, both morphologically and functionally. With these premises, numerous expectations were initially placed on anti-angiogenic therapies, soon dashed by their limited efficacy in concretely improving patient outcomes. Neovascularization in GBM soon emerged as a complex, dynamic, and heterogeneous process, hard to manage with the classical standard of care. Growing evidence has revealed the existence of numerous non-canonical strategies of angiogenesis, variously exploited by GBM to meet its ever-increasing metabolic demand and differently involved in tumor progression, recurrence, and escape from treatments. In this review, we provide an accurate description of each neovascularization mode encountered in GBM tumors to date, highlighting the molecular players and signaling cascades primarily involved. We also detail the key architectural and functional aspects characteristic of the GBM vascular compartment because of an intricate crosstalk between the different angiogenic networks. Additionally, we explore the repertoire of emerging therapies against GBM that are currently under study, concluding with a question: faced with such a challenging scenario, could combined therapies, tailored to the patient’s genetic signatures, represent an effective game changer?

Fusobacterium nucleatum (F. nucleatum), a Gram-negative anaerobe traditionally associated with periodontal disease, has recently emerged as a putative contributor to gastric carcinoma (GC) pathogenesis. Beyond its detection in gastric tissues, particularly in patients negative for Helicobacter pylori (H. pylori) or in advanced GC cases, F. nucleatum exerts diverse oncogenic effects. It promotes GC progression by modulating the tumor microenvironment through IL−17/NF-κB signaling, inducing tumor-associated neutrophils (TANs), upregulating PD-L1 expression, and enhancing immune evasion. Moreover, it increases tumor invasiveness via cytoskeletal reorganization, while extracellular vesicles (EVs) induced by the infection contribute to tumor cell proliferation, invasion, and migration. Clinically, its presence correlates with increased tumor mutational burden (TMB), venous thromboembolism, and poor prognosis. This review summarizes the current evidence regarding the emerging role of F. nucleatum in gastric tumorigenesis, examines its potential utility as a diagnostic and prognostic biomarker within the framework of precision oncology, and outlines the molecular methodologies presently employed for its detection in gastric tissue specimens.

In this article, we provide a review of large B-cell lymphomas (LBCLs), comparing the recently published fifth edition of the WHO classification and the International Consensus Classification (ICC) on hematolymphoid tumors. We focus on updates in the classification of LBCL, an heterogeneous group of malignancies with varying clinical behaviors and different pathological and molecular features, providing a comparison between the two classifications. Besides the well-recognized diagnostic role of clinical, morphological and immunohistochemical data, both classifications recognize the ever-growing impact of molecular data in the diagnostic work-up of some entities. The main aim is to offer a guide for clinicians and pathologists on how the new classifications can be applied to LBCL diagnosis in routine practice. In the first part of the paper, we review the following categories: LBLs transformed from indolent B-cell lymphomas, diffuse large B-cell lymphoma, not otherwise specified (DLBCL, NOS), double-hit/triple-hit lymphomas (DH/TH), high-grade large B-cell lymphoma, not otherwise specified (HGBCL, NOS), LBCL with IRF4 rearrangement, Burkitt lymphoma (BL) and HGBCL/LBCL with 11q aberration, focusing on the differences between the two classifications. In the second part of the paper, we provide a practical diagnostic algorithm when facing LBCLs in routine daily practice.

Upper tract urothelial carcinoma (UTUC) is a rare but aggressive malignancy with increasing incidence, often diagnosed at advanced stages due to the limitations of current diagnostic tools. Conventional methods such as urinary cytology, imaging, and ureteroscopy have important drawbacks, including low sensitivity, high costs, and procedural invasiveness. As a result, there is a growing need for non-invasive, highly accurate diagnostic approaches. Epigenetic biomarkers, particularly DNA methylation-based assays, have emerged as promising alternatives for UTUC detection and surveillance. Among these, Bladder EpiCheck® (BE) has shown remarkable sensitivity and specificity, particularly for high-grade tumors, making it a valuable adjunct to standard diagnostic techniques. By analyzing tumor-specific methylation patterns in urine samples, BE offers a practical and non-invasive solution that could improve early detection, reduce the need for ureteroscopy, and enhance risk stratification. Several studies have demonstrated its superior diagnostic accuracy, with sensitivity reaching 97.4 % and specificity up to 100 % for high-grade UTUC. Despite these advantages, challenges remain regarding the standardization of testing protocols, validation in larger patient cohorts, and evaluation of cost-effectiveness. Moreover, the role of DNA methylation biomarkers in guiding clinical decisions and predicting disease progression requires further investigation. This review explores the current state of UTUC diagnosis, compares BE with conventional and emerging biomarkers, and discusses its clinical applications, limitations, and future perspectives. The integration of molecular biomarkers like BE into clinical practice has the potential to revolutionize UTUC diagnosis, improving patient outcomes through more precise, non-invasive detection strategies.

Programmed death-ligand 1 (PD-L1) has become a central biomarker and therapeutic target across multiple solid tumors, yet its clinical meaning in prostate cancer (PCa) remains unsettled. PCa is commonly described as an immunologically ‘cold’ malignancy, characterized by limited baseline cytotoxic T-cell infiltration and a tumor microenvironment (TME) shaped by myeloid-driven suppression and low neoantigen load in many cases. Against this background, PD-L1 expression in PCa is typically low in untreated primary tumors but can increase in aggressive variants, advanced stages, and metastatic castration-resistant disease, where therapy pressure and microenvironmental cues may select for immune-evasive phenotypes. The literature is further complicated by major analytic variability, including differences in antibody clones and platforms, scoring algorithms (tumor proportion score, combined positive score, immune-cell scoring), cut-offs, tissue sites and timing, and pre-analytical variables such as fixation and decalcification. Collectively, available studies suggest that higher PD-L1 expression tends to be associated with adverse clinicopathological features and may enrich for responses to immune checkpoint inhibitors in selected settings, but PD-L1 immunohistochemistry alone is insufficient as a stand-alone predictive tool in unselected patients. This review synthesizes the biological drivers of PD-L1 regulation in PCa, dissects key methodological sources of heterogeneity in PD-L1 assessment, summarizes clinicopathological and therapeutic correlations, and outlines emerging biomarkers and approaches (including mismatch repair deficiency/microsatellite instability, tumor mutational burden, gene-expression signatures, liquid biopsies, and neuro-immune interactions) that may enable more actionable patient stratification.