The Journal of Pathology最新文献

Carcinosarcomas (CSs) are aggressive biphasic tumors characterized by epithelial and mesenchymal components, whose histogenesis and differentiation dynamics remain poorly understood. We present single-nucleus RNA sequencing (snRNA-seq) analysis of six CSs (five endometrial and one ovarian) and two normal endometrial samples, profiling over 96,298 cells. By integrating transcriptomic data with inferred copy number variations (CNVs), immunohistochemistry (IHC), fluorescence in situ hybridization (FISH), and in situ hybridization (ISH) validation, we resolved the complex cellular architecture of these tumors, identified lineage-specific programs, and revealed unexpected differentiation trajectories. snRNA-seq was used to further refine the histopathological classification of three cases by uncovering heterologous differentiation not previously recognized: one rhabdomyogenic, one osteogenic, and, notably, one exhibiting a novel tenogenic program, defined by the expression of SCX, MKX, and TNMD. All CSs displayed a prominent mesenchymal compartment comprising both undifferentiated fibroblast-like cells and distinct lineage committed populations, including rhabdomyoblasts (Rhab), tenoblasts (Teno), osteoblasts (Osteo), and chondroblasts (Chond). In some tumors, multiple mesenchymal identities co-existed, and in others, differentiation gradients (e.g. immature versus mature rhabdomyoblasts) were observed. These patterns underscore the cellular plasticity and multilineage potential of the sarcomatous component. Furthermore, the expression of specialized interface markers (COL22A1, NCAM1, ACAN, CHRNG, MUSK) suggests that some tumors use structured developmental programs reminiscent of the muscle–tendon junction, enthesis, or neuromuscular junction. CNV analysis revealed tumor-specific genomic alterations with clonal and subclonal patterns linked to differentiation state, which were validated by FISH. Altogether, this study demonstrates that CSs are not static biphasic tumors but rather complex ecosystems with extensive developmental plasticity. Our findings redefine their classification and support the use of single-nucleus approaches to uncover hidden differentiation trajectories in highly heterogeneous cancers, including the discovery of a previously unreported tenogenic lineage. Our results challenge the diagnosis of homologous CS when only morphological criteria are applied. © 2026 The Author(s). The Journal of Pathology published by John Wiley & Sons Ltd on behalf of The Pathological Society of Great Britain and Ireland.

BK polyomavirus (BKPyV) is a viral infection experienced by kidney transplant recipients that can lead to the development of BKPyV-associated nephropathy (BKPyVAN), graft dysfunction, and loss. There are no BKPyV-specific treatments available to prevent this significant cause of transplant failure. This bioinformatic study aims to characterise the cellular networks and pathways involved in BKPyVAN to identify novel therapeutic targets. Four publicly available bulk transcriptomic datasets containing BKPyVAN post-transplant biopsy tissue were identified in the National Centre for Biotechnology Information Gene Expression Omnibus (NCBI GEO). Differentially expressed genes (DEGs) (adjusted p < 0.05, fold change ≥ 1.5) were identified and dataset comparisons made between BKPyVAN versus stable grafts and acute rejection versus stable grafts. Canonical pathways were investigated using QIAGEN Ingenuity Pathway Analyses and protein interaction networks using STRING v12.0. There were 226 genes identified as differentially expressed in BKPyVAN compared with stable graft function that were conserved across all four datasets. This gene signature was associated with three cellular networks and 201 significantly enriched pathways. The cellular networks identified included 67 immune-related proteins; eight proteins associated with the absent in melanoma 2 (AIM2) inflammasome; and four HLA class II proteins. The most notable pathways significantly increased in BKPyVAN included HLA class II antigen presentation (p < 0.001), inflammasome (p < 0.001), and interleukin 6 signalling (p < 0.01). There were seven DEGs that were observed as common to all BKPyVAN versus acute rejection comparisons. The TUBB3 gene was the only gene that was consistently upregulated in all datasets. Several pathways and potential treatment targets were identified using a bulk RNA mining strategy, including HLA class II antigen presentation, AIM2 inflammasome, and IL-6 signalling in BKPyVAN pathology. Such tools provide an important first step in identifying novel mediators of disease pathogenesis and likely hold the key for the discovery of potential treatment targets for BKPyVAN in the future. © 2026 The Pathological Society of Great Britain and Ireland.

Tumor budding and poorly differentiated clusters are key prognostic indicators in colorectal cancer, yet the molecular mechanisms underlying their formation remain incompletely characterized. Using the GeoMx Digital Spatial Profiler, we analyzed gene expression profiles across tubular components, transitional zones, and tumor budding and poorly differentiated cluster regions in 12 colorectal cancer cases. While histopathological assessment revealed no definitive evidence of complete epithelial-mesenchymal transition in tumor budding or poorly differentiated clusters, transcriptional analysis demonstrated significant upregulation of epithelial-mesenchymal transition-related genes (ZEB1, ZEB2, SNAI2) alongside maintained epithelial marker expression, indicating a partial epithelial-mesenchymal transition phenotype. Tumor budding and poorly differentiated cluster regions showed reduced proliferative activity with significant downregulation of MYC and its target genes involved in protein synthesis and cell cycle progression. Conversely, RUNX2 and its targets were significantly upregulated in tumor budding and poorly differentiated cluster regions, particularly genes mediating cell adhesion, migration, and extracellular matrix interactions. KRT80 showed striking upregulation in tumor budding and poorly differentiated cluster regions, correlating with RUNX2 expression. This reciprocal pattern of MYC downregulation and RUNX2 upregulation appears to contribute to maintaining the hybrid epithelial–mesenchymal state characteristic of tumor budding and poorly differentiated cluster regions. Our findings reveal that the formation of these regions involves transcriptional changes preceding morphological alterations, with RUNX2 potentially driving this invasive phenotype while preserving epithelial features. These insights may enhance our understanding of the mechanisms driving colorectal cancer progression and identify potential therapeutic targets for aggressive disease. © 2026 The Pathological Society of Great Britain and Ireland.

Netherton syndrome (NS) is a rare, severe, and often life-threatening disease for which current therapeutic approaches are limited and show variable effectiveness. NS is characterized by excessive epidermal desquamation that results in a highly defective epidermal barrier, constitutive skin inflammation, allergies, and hair abnormalities. NS develops due to loss-of-function mutations in the SPINK5 gene, which encodes the LEKTI inhibitor that regulates KLK proteases (KLK5, KLK6, KLK7, KLK13, and KLK14). These findings indicate that dysregulation of proteolytic networks underlies the extensive skin shedding and inflammation characteristic of NS. Spink5^−/− mice recapitulate the major features of the human disease but exhibit neonatal lethality. Several double- and triple-knockout models have been generated to rescue the lethal NS phenotype, and have proved instrumental in studies aiming to elucidate the biological pathways involved in NS, and to identify and validate potential targets for drug development. These studies have established that inhibition of excessive KLK protease activity in LEKTI-deficient epidermis can reverse the cutaneous manifestations of NS. In particular, ablation of KLK5 results in a marked therapeutic response, although KLK7 or TNFα must also be inhibited to rescue the most severe (lethal) form of NS. Murine models have also been essential in proving or disproving putative pathways and/or therapeutic targets proposed from in vitro studies or patient case studies. Collectively, these models have provided a deeper understanding of the epidermal proteolytic cascades involved in NS pathology and in normal skin renewal. Moreover, these models offer a platform in which disease-specific candidate therapeutics can be tested and preclinically validated. © 2026 The Author(s). The Journal of Pathology published by John Wiley & Sons Ltd on behalf of The Pathological Society of Great Britain and Ireland.

A machine learning-based pathomics model was investigated for its value and biological significance in predicting overall survival (OS) after surgery in pancreatic cancer patients. Data from 173 patients with pancreatic ductal adenocarcinoma (PDAC) who underwent surgery and continued follow-up in two centers were retrospectively analyzed. Pathomics parameters of both the tumor and peritumor were measured in all patients, and the optimal pathomics score (Pathscore) was calculated using five machine learning methods. The best Pathscore was then combined with multiple clinical parameters to analyze its incremental value and to construct a comprehensive nomogram. TCGA data, multiplex immunofluorescence, spatial analysis, and single-cell sequencing were used to explore the biological mechanisms of pathomics. In predicting OS, pathomics parameters from the tumor and peritumoral regions provided complementary prognostic information. The LASSO-based combined model achieved the best predictive accuracy. Multivariate Cox regression analysis identified T-stage, N-stage, CA19-9, and Pathscore as independent predictors of OS in patients with PDAC. The integrated nomogram demonstrated superior and more stable predictive performance. Analysis of the TCGA dataset suggested that the pathomics model was associated with the immune status of pancreatic cancer, a finding supported by trends in the validation cohort. Spatial analysis and single-cell analysis further revealed a strong association between the Pathscore and immune cell infiltration, in particular CD8+ T cells. Machine learning-based pathomics models can help to predict the immune status and OS of patients with PDAC. The integration of pathomics with clinical parameters provides a robust basis for immune evaluation, prognostic prediction, and therapeutic decision-making in PDAC. © 2026 The Pathological Society of Great Britain and Ireland.

Neoadjuvant immunochemotherapy (NAIC) is a standard treatment for triple-negative breast cancer (TNBC), but there is no reliable biomarker to identify potential responders and optimize patient care. In this study, we developed a model named Immunotherapy Prediction based on Pathological Images (IPPI) by machine learning. The IPPI model performed well in the discovery cohort and two validation cohorts, which included a total of 209 patients, and its predictive power was significantly improved compared to clinical factors and the combined positive score for programmed death-ligand 1. TNBC patients predicted to achieve a pathological complete response had a better prognosis than those predicted to have residual disease. Moreover, we elucidated the relationship between histopathological features and biological characteristics, thereby improving the interpretability of the IPPI model. This study proposes a novel and efficient model to facilitate the prediction of NAIC response in TNBC patients, highlights key histopathological features associated with treatment response, and presents new evidence for precision immuno-oncology through the integration of machine learning and digital pathology. © 2026 The Pathological Society of Great Britain and Ireland.

Intratumor heterogeneity of ERBB2 amplification/HER2 overexpression is frequently observed in ERBB2-amplified high-grade endometrial carcinoma (HG-EC) and contributes to anti-HER2 therapy resistance. To elucidate the molecular pathogenesis and evolutionary trajectory of HER2-heterogeneous HG-ECs, we performed next-generation sequencing of spatially distinct HER2-negative (HER2−) and HER2-positive (HER2+) tumor areas from nine tumors (whole exome, n = 7; targeted panel, n = 2). HER2− and HER2+ components shared a high proportion of somatic mutations, particularly clonal mutations, including known EC driver genetic alterations. The 17q12 amplicon, containing the ERBB2 gene, was the only significant recurrent copy number alteration that differed between HER2− and HER2+ components. By unsupervised hierarchical clustering of genome-wide copy number alterations, samples clustered together at the patient level rather than by HER2 status. Intra- and intertumor heterogeneity in ERBB2 amplification/HER2 expression was also observed in metastatic lesions, which likely originated from different tumor subpopulations within the primary tumor. Exploratory spatial transcriptomics analyses revealed gene expression differences associated with HER2 status, including a shift from ‘mesenchymal-like’ toward epithelial differentiation in HER2+ components for a subset of cases, a finding that warrants further investigation. Our results suggest that HER2 heterogeneity in HG-EC reflects late acquisition of ERBB2 amplification during tumor evolution. ERBB2 does not appear to drive tumor initiation in HER2-heterogeneous HG-EC but likely serves a context-dependent role in the progression of established tumors. © 2026 The Pathological Society of Great Britain and Ireland.

Ovarian cancer is a heterogeneous gynecologic malignancy, with most cases diagnosed at an advanced stage. Despite the availability of effective treatments such as surgery and platinum/taxane-based chemotherapy, advanced ovarian cancer frequently recurs, highlighting the need to investigate mechanisms of chemotherapy resistance. Delta-like non-canonical Notch ligand 1 (DLK1), a transmembrane protein of the EGF-like family, is aberrantly expressed in several cancers. Our previous study demonstrated that DLK1 promotes oncogenic behaviors and epithelial–mesenchymal transition in ovarian high-grade serous carcinoma. In this study, we observed a positive correlation between DLK1 and stemness markers CD44 and CD133 in human epithelial ovarian cancer using tissue microarray analysis. Overexpression of DLK1 by an adenovirus vector accelerated sphere-forming capability and upregulated CD44, CD133, and ABCG2 expression in human ovarian cancer cell lines. Conversely, DLK1 silencing by siRNA abolished the stimulatory effects on the stem cell-like properties and reduced CD44 and CD133 expression in ovarian cancer cells. Furthermore, DLK overexpression by an adenovirus vector enhanced colony formation and suppressed the cisplatin- and taxol-induced death in human ovarian cancer cells. Conversely, DLK1 siRNA reversed cell death and colony formation induced by these chemotherapeutic agents. Finally, we demonstrated that DLK1 regulates CD44, CD133, and ABCG2 expression through the Notch1/AKT/STAT3 signaling pathway, involving the phosphorylation and translocation of STAT3. Collectively, these results suggest that targeting DLK1 may represent a potential therapeutic strategy to improve outcomes in recurrent ovarian cancer following chemotherapy. © 2026 The Pathological Society of Great Britain and Ireland.

Fumarate hydratase-deficient uterine leiomyomas (FHd-ULMs) represent a molecularly distinct subgroup of smooth muscle tumours associated with hereditary leiomyomatosis and renal cell cancer (HLRCC) syndrome. This study determined the detection rate of germline pathogenic FH variants in FHd-ULMs using paired tumour-normal sequencing and assessed the utility of integrated morphological, immunohistochemical (FH/2SC), genomic, and clinical features for selecting cases suspected of HLRCC. Histopathological assessment of the 252 FHd-ULMs revealed consistent morphological features, whereas molecular profiling identified three biologically distinct categories. Germline FH-mutated (hereditary) cases constituted 35.7% (74/207) of this FHd-ULM cohort, four of which showed concomitant somatic copy-number alterations. This contrasted with the somatic-mutated subgroup, which comprised 50.7% (105/207), including 18.4% (38/207) with isolated somatic copy-number losses. The remaining 13.6% (28/207), though lacking detectable FH mutations, were classified with somatic-mutated cases as sporadic FHd-ULMs based on shared clinicopathological features. Molecular analysis identified shared variant distributions across exons 2–10, with exons 5 and 7 in the fumarate lyase domain emerging as the predominant mutational hotspots. Notably, truncating mutations showed significantly higher prevalence in hereditary cases versus somatic variants (p < 0.01). Clinically, hereditary FHd-ULMs presented at younger ages (< 45 years) and manifested more aggressive phenotypes, including elevated rates of infertility (54.1% versus 26.4%), multifocal tumour development (86.5% versus 53.4%), increased surgical interventions (44.6% versus 11.5%), and familial leiomyoma clustering (51.4% versus 24.2%). Our results identify germline FH mutations, which confer significant HLRCC risk, in 35.7% of FHd-ULMs, while somatic alterations account for the majority of cases. To ensure efficient resource allocation, we propose a stratified diagnostic approach: initial universal FH/2SC immunohistochemical screening for ULMs displaying ≥ 3 FH-deficient morphological features, followed by confirmatory genetic testing in high-risk individuals, defined by either aberrant FH/2SC immunoreactivity or, in immunohistochemically normal cases, age < 45 years together with a personal/family history of multiple symptomatic leiomyomas or HLRCC-associated neoplasms. © 2026 The Pathological Society of Great Britain and Ireland.

Chronic kidney disease (CKD) is a major global health challenge affecting over 10% of the adult population. A hallmark of CKD progression is the transdifferentiation of kidney fibroblasts into extracellular matrix-producing myofibroblasts, a key mechanism involved in the decline of kidney function and the development of kidney failure. Fibroblasts maintain the structural integrity of the kidney and support epithelial survival, repair, and regeneration after acute kidney injury. Maladaptive repair is a failure to resolve fibroblast activation, which ultimately progresses to chronic injury and CKD. In this study, we showed that the membrane-bound coreceptor neuropilin 1 (NRP1) was essential to maintain fibroblast function and prevent their transdifferentiation into myofibroblasts. We used the myelin protein zero-Cre (P0-Cre) to specifically abrogate Nrp1 in kidney resident fibroblasts during fibrosis progression. We employed kidney-induced interstitial fibrosis models combined with a lineage-tracing strategy, single-cell RNA sequencing analysis, and ex vivo explant cultures to reveal a cell autonomous protective role for NRP1 in limiting fibrosis. Furthermore, we extended the analysis by showing that Nrp1 conditional mutants were more prone to develop cardiac fibrosis in a mouse model of heart failure. Collectively, these findings provide new insights into the signalling pathways controlling the transition from acute to chronic kidney disease conversion and identify NRP1 as a novel regulator of fibroblast supportive function. © 2025 The Author(s). The Journal of Pathology published by John Wiley & Sons Ltd on behalf of The Pathological Society of Great Britain and Ireland.