Pituitary neuroendocrine tumors (PitNETs) are common sellar neoplasms and represent a major component of routine pituitary pathology. In the 2022 World Health Organization (WHO) Classification, transcription factor-defined lineage assignment is central to diagnosis. However, lineage-related morphologic information on routine hematoxylin and eosin (H&E)-stained whole-slide images (WSIs) has not been systematically characterized. We developed an attention-guided graph neural network to predict PitNET lineage directly from H&E-stained WSIs and to identify regions prioritized for classification. Consecutive patients who underwent surgical resection for PitNETs at Beijing Tiantan Hospital between 2021 and 2025 were included. In the internal hold-out set, the five fold-specific models achieved a mean F1-score of 92.78% and a mean balanced accuracy of 94.84%. In a temporally independent validation cohort, the final model achieved an F1-score of 87.64% and a balanced accuracy of 89.48%. Performance varied across lineages and histomorphologic subtypes, and the most common error pattern was misassignment of PIT-1 and T-PIT tumors to SF-1. Attention maps predominantly highlighted tumor-rich regions. Quantitative cell-level morphometry supported lineage-associated patterns, including larger cell and nuclear sizes in PIT-1 tumors, more elongated nuclei in SF-1 tumors, and higher cellular density with reduced intercellular spacing in T-PIT tumors. In six cases with multiple synchronous PitNETs of distinct lineages, patch-level prediction maps corresponded closely to transcription factor immunohistochemistry. In a small exploratory subset with available DNA methylation data, methylation classifier results were concordant with model predictions in 11 of 15 cases and with routine histopathologic diagnoses in 8 of 15 cases. These findings indicate that routine H&E-stained WSIs contain learnable morphologic information related to PitNET lineage. Attention-guided spatial modeling provides an interpretable framework for characterizing lineage-associated patterns in PitNETs.

Adrenal lesions often occur in patients with multiple endocrine neoplasia type 1 (MEN1), mostly adrenal cortical adenomas (ACAs), although the frequency of adrenal cortical carcinomas (ACCs) is higher than in the general population. The coexistence of benign and malignant masses has seldom been documented, leaving open the question of ACC progression from benign forms. We report a comprehensive genetic characterization of three adrenal cortical tumor samples obtained from a familial MEN1 patient, operated for the rapid progression of an initially stable nonfunctional adrenal incidentaloma. Histologically, the tumor consisted of a small ACA contiguous to a large ACC, which subsequently relapsed. Exome sequencing of ACC, ACA and recurrence evidenced a MEN1 loss of heterozygosity (LOH) in ACC but not in ACA, where, however, a second hit driven by alternative mechanisms could not be excluded. The majority of the ACA variants were found to co-occurred in ACC (n = 36/42) and were benign, except for two of unknown significance in KANK1 and REN genes, described as associated with renal cancer. Among variants shared between ACC and its recurrence (n = 69), 11 were Tier III, while 2 affecting TP53 and NF1 genes were pathogenic. Bioinformatic clonal evolution analysis identified one clone - characterized by TP53 and NF1 mutations - absent in ACA but present in ACC and recurrence, as well as 2 clones shared between ACA and ACC but lost in the recurrence. In conclusion, comparative Whole Exome Sequencing (WES) analysis of three adrenal tumors in a MEN1 patient suggests a possible relationship between malignant and benign lesions occurring in MEN1 patients, without, however, demonstrating any causal adenoma-to-carcinoma progression driven by MEN1 LOH. Overall, these data further suggest an increased risk of MEN1 patients to develop adrenocortical malignancy.

Rb and p53 status helps distinguishing well from poorly differentiated high-grade neuroendocrine neoplasms (HG-NENs) and may predict response to platinum-based chemotherapy. Historically, gene and protein profile correlations for Rb and p53 were performed on frozen samples using high-throughput techniques efficient for copy number determination. This situation differs from routine practice in which NGS is not available everywhere and small deletions may be hard to detect on formalin-fixed paraffin-embedded (FFPE) samples with a gene panel approach. This study aimed at describing the immunohistochemical (IHC) staining patterns of Rb, p16, and p53 in a large cohort of well and poorly differentiated lung and pancreatic HG-NENs, and comparing the performance of Rb/p53 NGS profiling with IHC to detect molecular alterations. Rb, p16, and p53 protein status were determined by IHC in 132 HG-NENs (79 pulmonary and 53 pancreatic). NGS was performed on a subset of 63 tumors. Three IHC patterns were observed in HG-NENs: (i) loss of expression (Rbloss = 45.4% with p16strong expression in most cases, p53loss = 32.6%), (ii) variable expression (Rbvar = 40.8% including 12 cases with p16strong, p53var = 37.1%), (iii) strong expression (Rbstrong = 13.6%, p53strong = 30.3%). NGS testing and IHC agreed in 76.2% and 82.5% of cases for Rb and p53, respectively. Abnormal Rb & p53 IHC profiles, highly suggestive of a molecular alteration, were missed by NGS in 42% and 20.5% of cases, respectively. In conclusion, IHC is a reliable, rapid, and cost-effective screening method in HG-NENs for Rb/p53 status.

Pheochromocytomas (PCCs) and paragangliomas (PGLs, together PPGLs) are the most hereditary tumors known. PPGLs were considered benign, but the fourth edition of the World Health Organisation (WHO) classification redefined all PPGLs as malignant neoplasms with variable metastatic potential. The metastatic rate differs based on histopathology, genetic background, size, and location of the tumor. The challenge in predicting metastatic disease lies in the absence of a clear genotype–phenotype correlation among the more than 20 identified genetic driver variants. Recent advances in molecular clustering based on underlying genetic alterations have paved the way for improved cluster-specific personalized treatments. However, despite some clusters demonstrating a higher propensity for metastatic disease, cluster-specific therapies have not yet been widely adopted in clinical practice. Comprehensive genomic profiling and transcriptomic analyses of large PPGL cohorts have identified potential new biomarkers that may influence metastatic potential. It appears that no single biomarker alone can reliably predict metastatic risk; instead, a combination of these biomarkers may be necessary to develop an effective prediction model for metastatic disease. This review evaluates current guidelines and recent genomic and transcriptomic findings, with the aim of accurately identifying novel biomarkers that could contribute to a predictive model for mPPGLs, thereby enhancing patient care and outcomes.

Pituitary neuroendocrine tumors (PitNET) that metastasize comprise ~ 0.2% of adenohypophyseal tumors are aggressive and are challenging to treat. However, many non-metastatic tumors are also aggressive. Herein, we review 21 specimens from 13 patients at UCSF with metastatic PitNETs (CSF or systemic, N = 7 patients), high-grade pituitary neuroendocrine neoplasms (HG-PitNEN, N = 4 patients), and/or PitNETs with sarcomatous transformation (PitNET-ST, N = 5 patients). We subtyped cases using the World Health Organization (WHO) and International Agency for Research on Cancer (IARC) criteria for neuroendocrine neoplasms (NENs). Lineage subtypes included acidophil stem cell, null cell, thyrotroph, corticotroph, lactotroph, and gonadotroph tumors. The median Ki-67 labeling index was 25% (range 5–70%). Lack of p16 was seen in 3 cases, with overexpression in 2. Strong diffuse p53 immunopositivity was present in 3 specimens from 2 patients. Loss of Rb expression was seen in 2 cases, with ATRX loss in one. Molecular analysis in 4 tumors variably revealed TERT alterations, homozygous CDKN2A deletion, aneuploidy, and mutations in PTEN, TP53, PDGFRB, and/or PIK3CA. Eight patients (62%) died of disease, 4 were alive at the last follow-up, and 1 was lost to the follow-up. All primary tumors had worrisome features, including aggressive lineage subtype, high mitotic count, and/or high Ki-67 indices. Additional evidence of high-grade progression included immunohistochemical loss of neuroendocrine, transcription factor, and/or hormone markers. We conclude that metastatic PitNET is not the only high-grade form of pituitary NEN. If further confirmed, these histopathologic and/or molecular features could provide advanced warning of biological aggressiveness and be applied towards a future grading scheme.

Little is known about the morphomolecular features of G3 neuroendocrine tumors (G3NETs) under prolonged systemic treatments, although rapid progression is increasingly observed. This longitudinal study aims to elucidate the course and morphomolecular features of metastasized G3NETs with high-grade transformation. Clinical and histological findings in 40 patients with metastasized and treated G3NETs, which were histologically examined at least twice with an interval time of more than 6 months (median 27), were reviewed and the morphomolecular changes recorded and assigned to treatment. Neuroendocrine carcinoma (NEC)-like histology defined by high-grade atypia, diffuse growth pattern, and/or necrosis was identified in nine (22%) G3NETs (seven pancreatic, two rectal) patients. All NEC-like tumors showed a significantly higher Ki67 increase and longer interval time between first and last examination than non-NEC-like G3NETs (53 vs. 19% and 60 vs. 24 months, respectively). Moreover, all NEC-like G3NETs had TP53 (100%), but rarely RB1 (12%) mutations, and retained NET-typical mutations such as MEN1 or DAXX (five of the pancreatic NETs). The last treatments received prior to the NEC-like transformation included PRRT (n = 3), somatostatin analog, everolimus, sunitinib (n = 1 each), and alkylating agents (n = 2). Abrupt clinical progression in patients with metastasized G3NETs is associated with a significant increase in Ki67, accelerated growth, and NEC-like histology. These findings are most likely attributable to the novel TP53 mutation, which was detected in all nine cases at the last evaluation. However, none of the cases exhibited a complete transformation to a typical NEC, as the tumors retained partial histological and genetic features of NETs.

Glucagonomas are functioning pancreatic neuroendocrine tumors (PanNETs) responsible for glucagonoma syndrome. This study aims to shed light on the clinicopathological and molecular features of these neoplasms. Six patients with glucagonomas were identified. All neoplasms were investigated with immunohistochemistry for neuroendocrine markers (Synaptophysin, Chromogranin-A), ATRX, DAXX, ARX, and PDX1 transcription factors. Fluorescent in situ hybridization (FISH) for assessing alternative lengthening of telomeres (ALT), and next-generation sequencing (NGS) for molecular profiling were performed. All cases were large single masses (mean size of 8.2 cm), with necrolytic migratory erythema as the most common symptom (6/6 cases, 100%). All neoplasms were well-differentiated G1 tumors, except one case that was G2. The tumors consistently showed classic/conventional histomorphology, with solid-trabecular and nested architecture. Lymphatic and vascular invasion (6/6, 100%), perineural infiltration (4/6, 66.6%), and nodal metastasis (4/6, 66.6%) were frequently observed. Transcription factors expression showed strong ARX expression in all tumors, and PDX1 expression in 5/6 cases (83.3%), indicating co-occurring alpha- and beta-cell differentiation. NGS showed recurrent somatic MEN1 and ATRX/DAXX biallelic inactivation. Cases with ATRX or DAXX mutations also showed matched loss of ATRX or DAXX protein expression and ALT. One case harbored somatic MUTYH inactivation and showed a high tumor mutational burden (TMB, 41.0 mut/Mb). During follow-up, one patient died of the disease, and four patients developed distant metastasis. Pancreatic glucagonomas are distinct PanNETs with specific clinicopathological and molecular features, including histological aspects of biological aggressiveness, co-occurring alpha- and beta-cell differentiation, MEN1 and DAXX/ATRX mutations enrichment, and the possible presence of high-TMB as an actionable marker.

A 45-year-old male presented with acute mastoiditis and facial nerve paralysis.Computerized tomography of the temporal bone demonstrated complete opacification of left mastoid air cells, near complete middle ear cavity opacification, and bony erosion of the petrous apex extending to the left geniculate ganglion and proximal facial nerve canal.There was associated tympanic membrane thickening and external auditory canal debris.The patient was referred for mastoidectomy at which point a biopsy was performed.

Uno de los principales escollos de los estudiantes de espanol como lengua extranjera de nivel intermedio y avanzado es el aprendizaje de las colocaciones. Algunos investigadores de la teoria sentido-texto han propuesto servirse de las funciones lexicas –una herramienta lexicografica para que el estudiante aprenda las relaciones lexicas–, pero hasta ahora se han desarrollado pocos trabajos. En este articulo intentamos mostrar como la utilidad de los sentidos genericos de dichas funciones, llamados «glosas», depende de su sistematizacion y proponemos su uso desde los primeros estadios del aprendizaje. Se invita a los profesores de ELE a desarrollar materiales sirviendose de dichos sentidos y se proponen, a modo de ejemplo, ejercicios de colocaciones para los distintos niveles de ELE. Palabras clave: espanol como lengua extranjera (ELE), colocaciones, teoria sentido-texto, funciones lexicas, parafrasis y glosas. One of the major obstacles when learning Spanish as a foreign language (ELE) at intermediate and advanced level is the study of collocations. Some Meaning-Text Theory researchers have suggested using lexical functions – a lexicographical tool to help students learn lexical relationships – but very little work has been carried out to date. In this paper we claim that a well-structured set of meanings of these lexical functions, called «glosses», can be useful in ELE and show they can even be used at lower levels of ELE. We give some examples of the learning activities and exercises according to the ELE level, and invite teachers to develop materials according to this framework.