Hepatocellular carcinoma (HCC) is the most prevalent subtype of primary liver cancer. Immunotherapy, particularly targeting immune checkpoints such as programmed cell death protein 1 (PD-1), has shown considerable therapeutic promise. Bioinformatic analysis of Gene Expression Omnibus datasets demonstrated significant upregulation of WNK lysine deficient protein kinase 4 (WNK4) expression in HCC tissues obtained from untreated patients or anti-PD-1 non-responders, suggesting a potential role for WNK4 in HCC progression and immunotherapy resistance. Through a series of experiments, we confirmed that WNK4 promoted the proliferative and migrative abilities of HCC cells and enhanced the resistance of HCC model mice to anti-PD-1 therapy. The tumor microenvironment, particularly cancer-associated fibroblasts (CAFs), critically influences immunotherapy efficacy. The current study uncovered that WNK4 was transferred by HCC cell-derived exosomes into CAFs and promoted the cysteine metabolic reprogramming. Moreover, WNK4-mediated promotional effects of CAFs on the malignant phenotypes of HCC cells and anti-PD-1 resistance were in a cysteine-dependent manner. According to mechanism investigation, WNK4 could bind high mobility group box 1 (HMGB1) and induce its phosphorylation and cytoplasmic retention, thus reducing nuclear HMGB1-p53 interaction to enhance cystathionine gamma-lyase (CTH) expression. In summary, this study unveils a novel WNK4-HMGB1-p53 axis in CAFs that promotes HCC progression and modulates cysteine metabolism to foster immunotherapy resistance, offering potential therapeutic targets for HCC.

Ovarian cancer remains a major global health concern and leading cause of mortality among women due to late diagnosis, therapeutic resistance, and limited predictive biomarkers for treatment response. There is an urgent need for integrative approaches to improve early detection and treatment outcomes. In this study, we integrated machine learning and pharmacogenomics to identify drug-sensitive biomarkers and prioritize therapeutic candidates in ovarian cancer. Pharmacogenomic data were obtained from the Cancer Cell Line Encyclopaedia (CCLE) and Genomics of Drug Sensitivity in Cancer (GDSC-v2), including gene expression and drug response profiles across ovarian cancer cell lines. Predictive models were developed for six FDA-approved drugs using Elastic Net, Ridge, and Lasso regression, demonstrating robust predictive performance achieving Pearson correlation (r) to 0.65 and Spearman correlation (ρ) to 0.63 on validation sets. Biomarker analysis identified key genes associated with drug response, including CCR10 and PLEKHH2, WNT9B, ITPRID1, CHRNG, and DIRAS3. Ligand-based similarity screening against the COCONUT database followed by molecular docking and MD simulation identified three promising compounds (662142, 733302, and 883576) with improved binding affinity and conserved interactions with Topoisomerase-1. This integrative framework highlights the potential of combining machine learning, pharmacogenomics, and molecular modelling for biomarker discovery and drug prioritization in ovarian cancer.

Urothelial carcinoma (UC) is a common malignancy; however, UC subtypes/divergent differentiation (S/DD) accounts for only 10%-20% of overall cases. S/DD's aggressive biological behavior significantly affects its prognosis and therapeutic decision-making; thus, elucidating its genomic landscape within UC is important. This retrospective observational study aimed to evaluate and compare the molecular characteristics of S/DD and pure urothelial carcinoma (PUC).

Testing patients with non-small cell lung cancer for actionable variants is essential for guiding treatment decisions in accordance with established cancer care guidelines, though limited quantity and quality of tumor tissue often leaves insufficient material for comprehensive testing. Cytopathology specimens obtained through minimally invasive techniques are a potential source of diagnostic material for genomic profiling, though typically challenging to analyze.

Black women are more likely to develop and die from early-onset breast cancer. Hereditary breast cancer (HBC) is strongly associated with early age of onset. Accordingly, HBC characteristics in young Black women were evaluated.

Ovarian cancer peritoneal metastasis remains a major cause of recurrence and death despite advances in cytoreductive surgery, platinum-based chemotherapy, PARP inhibition, and immune checkpoint blockade. The limited activity of immunotherapy in this setting reflects layered immune resistance shaped by impaired antigen visibility, redundant inhibitory receptor networks, suppressive myeloid and regulatory circuits, and metabolic-epigenetic constraints within ascites and multicellular spheroids. These compartment-specific features distinguish peritoneal disease from anatomically confined tumors and help explain why systemic immune reinvigoration alone rarely produces durable benefit. Here, we synthesize current evidence on the mechanisms that govern immune escape in ovarian cancer peritoneal dissemination, with emphasis on antigen presentation defects, checkpoint-driven T-cell exhaustion, anti-phagocytic signaling, soluble suppressive mediators, and metabolic remodeling of the ascites microenvironment. We further examine how DNA damage response states intersect with innate immune sensing and discuss the translational implications of homologous recombination deficiency for combination treatment design. Finally, we propose a biomarker-guided framework that links antigen-presentation competence, immune engagement, dominant suppressive axes, and ascites-specific biology to rational therapeutic matching. This mechanism-centered view supports more precise trial design and provides a roadmap for combination immunotherapy in advanced ovarian cancer.

Guanine nucleotide-binding protein gamma 10 (GNG10) is implicated in various biological processes, yet its specific oncogenic role in colorectal cancer (CRC) remains poorly defined. This study aimed to elucidate the expression patterns, biological functions, and underlying mechanisms of GNG10 in CRC progression. We integrated TCGA datasets with tissue microarray immunohistochemistry and multivariate Cox regression models to evaluate the clinical significance of GNG10. Functional impacts on CRC malignant phenotypes and cancer stemness were assessed through gain- and loss-of-function models in vitro and in vivo. Mechanistic insights were gained via GSEA, Western blotting, and dual rescue strategies employing both pharmacological inhibition (Box5) and genetic depletion (shRHOA). We found that GNG10 was markedly overexpressed in CRC tissues, correlating with advanced pathological stage and poor overall survival. Multivariate analysis indicated that the prognostic value of GNG10 is closely associated with tumour progression. Functionally, GNG10 knockdown inhibited CRC cell proliferation, migration, and stemness while promoting apoptosis. Mechanistically, GNG10 activated the non-canonical Wnt/RHOA/JNK/NFATc1 signalling axis. Crucially, manipulation of GNG10 did not affect active or total β-catenin levels, thereby excluding canonical Wnt involvement. Both pharmacological inhibition with Box5 and genetic ablation of RHOA effectively abrogated GNG10-induced oncogenic phenotypes and the upregulation of cancer stem cell markers (CD44, CD133, OCT4, Nanog, SOX2). In vivo xenograft models confirmed that GNG10 knockdown suppressed tumour growth and decreased the expression of proliferation and stemness markers. Our findings demonstrate that GNG10 promotes CRC progression and stemness via the non-canonical Wnt signalling pathway. These findings highlight GNG10 as a promising prognostic indicator and a vulnerable target in CRC.

HER2 (ERBB2) gene mutations are associated with poor prognosis in non-small cell lung cancer (NSCLC). It is important to identify patients with HER2-mutant (HER2m) NSCLC, as new targeted treatment options are emerging. However, HER2 testing in clinical practice varies across regions, and real-world data on HER2m NSCLC are limited.

Neuroendocrine tumours (NETs) are well-differentiated epithelial neuroendocrine neoplasms that frequently develop in the small intestine, pancreas, and lungs. NETs originate from neuroendocrine cells specialized in hormone secretion implicated in a number of physiological processes. Their malignant transformation is characterized by low mutational burden, suggesting that epigenetic mechanisms may be at play. Recent understanding of epigenetic events driving cancer cell plasticity and tumour initiation has led to advances in the diagnosis and prognosis of NETs. Here, we provide a brief overview of NETs, including their current diagnosis and management, and present recent progress in understanding the role of epigenetic regulation, highlighting how this may influence NET tumorigenesis and may be used in therapeutic applications. Finally, this literature review emphasizes the need to gather more data on these rare malignancies to improve patient outcome.

Objective: To analyze the occurrence and distribution of high-level microsatellite instability (MSI-H) in precancerous lesions and histological subtypes of endometrial cancer, to compare the concordance rate between mismatch repair protein deficiency (dMMR) and MSI-H, to investigate the incidence of minimal microsatellite shift in the development of endometrial cancer, and to explore potential solutions for accurate MSI-H diagnosis. Methods: A total of 848 endometrial lesion samples that underwent molecular typing at the Department of Pathology, Obstetrics and Gynecology Hospital of Fudan University from January 2023 to January 2024 were collected, including 93 cases of atypical endometrial hyperplasia (AEH) and 755 cases of endometrial carcinoma. Microsatellite status, MMR protein expression, and gene mutations (MMR, POLE, etc.) were analyzed by histology, IHC, and NGS. Microsatellite status was determined by calculating the MSI-score based on 34 microsatellite loci. An MSI-score<0.3 was defined as microsatellite stable (MSS),≥0.3 as microsatellite instability-high (MSI-H), and 0.3±0.05 as the equivocal range. Simultaneously, a visualization graph was established for minimal microsatellite shift analysis and identification. Results: The 848 lesions included 93 cases of AEH, 442 cases of low-grade endometrioid carcinoma (G1-G2), 143 cases of high-grade endometrioid carcinoma (G3), 20 cases of clear cell carcinoma, 20 cases of carcinosarcoma, 16 cases of mixed adenocarcinoma, 12 cases of dedifferentiated/undifferentiated carcinoma, and 102 cases of other types (including 94 serous carcinomas, 6 mesonephric-like adenocarcinomas, and 2 gastric-type mucinous adenocarcinomas). The results showed that the incidence of dMMR in AEH was 4.3%, and MSI-H was 3.2%, significantly lower than the 24.9% and 23.0% in endometrial cancer, respectively (P<0.001). All the 102 cases of other types, including serous carcinoma, were pMMR/MSS. The overall concordance rate between dMMR and MSI-H was 90.6%, but it varied between 75% and 100% across different stages and histological subtypes. In molecular subtyping, the concordance rate was 75% for AEH and carcinosarcoma, 87.5% for dedifferentiated/undifferentiated carcinoma, 91.5% for high-grade endometrioid carcinoma, and 100% for other high-grade carcinomas, while the POLE-mutated subtype had the lowest concordance rate of 66.7%, significantly lower than the 93.1% overall concordance rate for the MSI-H subtype (P<0.001). In MSI-H cases, up to 84.5% of endometrial cancer cases exhibited minimal microsatellite shift in at least one locus, with 67.8% showing shifts in≥3 loci. Through analysis of 34 microsatellite loci and visualization, 20 cases (11.5% of 174 MSI-H cases) were identified as borderline (MSI-score=0.3±0.05), considered diagnostically challenging. Among these, MLH1-/PMS2- co-loss accounted for 12/20 cases, with half (6/12) harboring MLH1 mutations; isolated MSH6 loss accounted for 6/20 cases, with 5 of these harboring MSH6 mutations. The 20 minimal shift MSI-H cases shared all pathological features of typical MSI-H endometrial cancer. Conclusions: The incidence and distribution of MSI-H show significant differences across histological subtypes between endometrial precancerous lesions and endometrial carcinomas. The overall concordance between dMMR and MSI-H is good, but varies across different disease stages, histological types, and molecular subtypes. Minimal microsatellite shift is commonly detected in MSI-H cases, and some cases are difficult to interpret due to their classification within the equivocal range. Increasing the number of microsatellite loci, combined with visualization graph comparison and integration of mismatch repair protein immunophenotype and histological features, can effectively improve the accuracy of MSI-H interpretation.

Objective: To investigate the clinicopathological and molecular genetic characteristics of uterine inflammatory myofibroblastic tumor (UIMT). Methods: A retrospective analysis was conducted on 8 patients diagnosed with UIMT at Changzhou First People's Hospital, Changzhou Wujin Hospital of Traditional Chinese Medicine, Changzhou Second People's Hospital, and Changzhou Maternal and Child Health Care Hospital from January 2018 to July 2025, including 6 cases of non-aggressive UIMT, 1 case of aggressive UIMT and 1 case of epithelioid inflammatory myofibroblastic sarcoma (EIMS). The clinicopathological features, molecular genetic characteristics, treatment and prognosis were summarized. Results: The patient's age was 44.5(38.5, 46.0) years old, and the tumor size was 6.35(5.05, 8.70) cm. Among the cases, one tumor was located submucosally within the uterus, four were within the uterine muscular wall, and three were subserosa. The tumors exhibited heterogeneous morphological features: non-aggressive UIMT displayed either a mucinous pattern with sparse cellularity or a dense fascicular pattern (similar to smooth muscle), with well-defined tumor margins or invasive growth; UIMT with aggressive behavior showed invasive growth, with greater cellular atypia, more mitotic figures, and necrosis. One EIMS showed invasive growth, mainly composed of epithelioid cells with focal spindle cell morphology, significant cellular atypia, active mitotic figures, and a small amount of lymphocyte and plasma cell infiltration between the epithelioid cells. Fluorescence in situ hybridization (FISH) testing detected ALK rearrangement in all eight cases. RNA next generation sequencing (NGS) indicated that all tumors were driven by ALK rearrangement. Among these tumors, EIMS harbored a novel fusion gene, KANK2::ALK. DNA NGS demonstrated that UIMT with aggressive behavior and EIMS harbored more pathogenic gene alterations, including mutations in TP53 and ATM, as well as amplification of the CCNE1 gene. Except for EIMS, immunohistochemical analysis revealed positive expression of ALK protein in 7 cases of UIMT. Immunohistochemistry also showed abnormal expression of p16 or p53 protein in EIMS and one case of UIMT with aggressive behavior. All patients underwent surgical treatment. The follow-up period ranged from 4 to 84 months. One patient with aggressive behavior of UIMT developed liver and lung metastases 4 months after the surgery. The patient was treated with the oral targeted drug crizotinib and died of multiple organ failure 18 months after surgery. Conclusions: UIMT is a morphologically heterogeneous neoplasm, and its biological behavior can vary from benign to highly aggressive. This type of tumor is mainly driven by ALK gene rearrangement. UIMT and EIMS that exhibit aggressive behavior typically possess a greater number of genetic alterations. The abnormal expression of p53 or p16 protein, when combined with clinicopathological parameters, can serve as indicators for predicting the adverse biological behavior of tumors.

Objective: To investigate the clinicopathological features of tuberous sclerosis complex (TSC)-associated multifocal micronodular pneumocyte hyperplasia (MMPH) and the mutation status of TSC1/TSC2 genes. Methods: A retrospective analysis was conducted on 8 MMPH cases diagnosed at Shanghai Pulmonary Hospital Affiliated to Tongji University from September 2020 to August 2024. Clinical information and imaging findings of these cases were collected. Histopathological analysis and next-generation sequencing (NGS) were performed. Results: Among the 8 patients, there were 4 males and 4 females, aged 44 (36, 50) years old. None of them had a definite family history of TSC. Only 1 patient presented with respiratory symptoms at the time of consultation. Four patients had TSC-related cutaneous lesions. Two had shagreen patches, 1 had hypomelanotic macules, and 1 had hypomelanotic macules combined with fibrous plaques on the scalp and angiofibromas. Three patients themselves or their first-degree relatives had a history of epilepsy. Five patients themselves or their first-degree relatives had liver/kidney cysts or a surgical history of renal angiomyolipoma. Computerized tomography scans showed multiple ground-glass nodules in both lungs, with an average diameter of 10.5 (8.5, 11.0) mm. Six cases were diagnosed via surgical resection, and 2 via transbronchial cryobiopsy. Intraoperative frozen sections of the 6 surgically resected cases were all misdiagnosed as early-stage lung adenocarcinoma (including adenocarcinoma in situ, minimally invasive adenocarcinoma, and invasive non-mucinous adenocarcinoma), but the diagnosis was corrected to MMPH postoperatively. One of these 6 cases was concurrent with a benign perivascular epithelioid cell tumor (PEComa). Microscopically, MMPH was characterized by multifocal proliferation of type Ⅱ alveolar epithelial cells, which showed bland cellular morphology without atypia or invasive growth. For the case concurrent with PEComa, clear epithelioid cells were observed growing in nests or sheets. Immunohistochemically, the proliferative epithelial cells were positive for TTF-1 and showed low expression of Ki-67. PEComa cells were positive for Melan A and MiTF. NGS showed that TSC1 mutations were detected in 6 cases, TSC2 mutation in 1 case (which also had a concurrent class Ⅲ missense mutation of BRAF G466E), and no mutation in 1 case, with an overall mutation detection ratio is 7/8. No other lung cancer-related driver-gene variations were found in any case. Follow-up data were available for 8 patients. During the follow-up, 1 patient developed scattered thin-walled lucencies in both lungs, but no pathological examination was performed. All patients had a favorable overall survival status, follow-up time was 14.0(9.5,47.0) months. Conclusions: MMPH is a rare, benign TSC-associated pulmonary lesion, often presenting as multiple ground-glass nodules in both lungs. It can be easily misdiagnosed as early-stage lung adenocarcinoma. The diagnosis requires a comprehensive judgment based on clinical data, imaging findings, histopathology, and TSC1/TSC2 gene mutation results.

Objective: To investigate the diagnostic value of PRAME immunohistochemistry in the differential diagnosis between clear cell sarcoma (CCS) of soft tissue and melanoma, and its utility in routine diagnostic practice. Methods: A retrospective analysis was conducted on 25 CCS and 25 site-matched melanoma cases diagnosed at the Fudan University Shanghai Cancer Center from 2021 to 2025. Clinicopathologic data were collected, histologic features were assessed, and PRAME immunohistochemistry was performed. Clinicopathologic differences between the two groups were compared. The sensitivity and specificity of PRAME for differential diagnosis were evaluated. Results: Twenty-five patients with CCS were included, aged 38.0 (31.5, 56.0) years old. Another 25 patients with melanoma were included, aged 61.0 (54.0, 72.0) years old. In each group, 11 male and 14 female patients were included. Tumors were predominantly located in the extremities and trunks in both groups. In primary or recurrent melanomas, epidermal involvement with pagetoid proliferation at the dermoepidermal junction was frequently observed, whereas epidermal involvement was usually absent in CCS. Epithelioid morphology with prominent tumor-infiltrating lymphocytes and scant fibrous collagen septa were more often observed in metastatic melanomas, whereas a mixed epithelioid-spindle pattern, paucity of tumor-infiltrating lymphocytes, and frequent fibrous collagen septa were more commonly observed in CCS. PRAME negativity was observed in 96.0% (24/25) of the CCS cases, with only one case showing 5% nuclear positivity. By contrast, PRAME positivity was observed in 88.0% (22/25) of the melanoma cases, including diffuse nuclear positivity in>50% of tumor cells in 80.0% (20/25) of cases (P<0.001). With the cut-off of >50% positive cells, the sensitivity and specificity of PRAME for the diagnosis of melanoma were 80.0% and 100.0%, respectively. BRAF V600E immunohistochemistry was positive only in melanoma (56.0%, 14/25). All CCS cases were molecularly confirmed to harbor EWSR1 rearrangement.No co-occurrence of EWSR1 rearrangement and BRAF V600E immunohistochemistry positivity was observed in this cohort. Conclusions: Diffuse PRAME nuclear positivity is highly specific for diagnosing melanoma. It is particularly useful in biopsy specimens from recurrent lesions, needle biopsy specimens from metastatic lesions with inconspicuous pigmentation, biopsies of small primary lesions, and ulcerated specimens lacking an epidermal component. Therefore, PRAME immunohistochemistry is especially practical for community-hospital laboratories and interpreting biopsy specimens with limited tissue.

Objective: To investigate the clinicopathological characteristics of IDH-mutant adult-type diffuse gliomas. Methods: Adult-type diffuse gliomas with IDH1 (R132) or IDH2 (R172) mutation confirmed by Sanger sequencing and 1p/19q FISH testing were collected at the West China Hospital of Sichuan University between January 2021 and June 2025. The correlations among patient sex, age, lesion location, IDH mutation type, tumor type, and WHO grade were retrospectively analyzed. Results: Among the 1 301 patients, 722 (55.5%) were male and 579 (44.5%) were female, with a male-to-female ratio of 1.25∶1.00. The patient age was 41 (33, 51) years. There were 9 patients (0.7%) aged≤18 years, 1 140 (87.6%) aged 19-55 years, and 152 (11.7%) aged>55 years. There were observed in 1 271 cases (97.7%) of supratentorial tumor and 17 cases (1.3%) of infratentorial tumor, while the location was unknown in 13 cases (1.0%). IDH1 (R132) mutations were identified in 1 244 cases, with R132H being the most common (1 211 cases, 93.1%), while non-R132H IDH1 mutations were detected in 33 cases (2.5%) and IDH2 mutations in 57 cases (4.4%). Among the 1 301 cases, 711 cases (54.7%) were astrocytoma, with WHO grades 2, 3, and 4 comprising 364 cases (51.2%), 116 cases (16.3%), and 231 cases (32.5%), respectively. Oligodendrogliomas accounted for 590 cases (45.3%), with WHO grade 2 in 372 cases (63.1%), and WHO grade 3 in 218 cases (36.9%). Patient sex was associated with WHO grade (r=0.078, P=0.014), but not age, location, tumor type, or IDH mutation type. Male patients outnumbered females across all grades, with the proportion of males increasing with higher tumor grades. Patient age was associated with IDH mutation type and tumor type (r=-0.072, 0.199, P=0.001), but not location or WHO grade. Compared to patients>18 years, those≤18 years were more likely to harbor non-R132H IDH1 mutations, whereas IDH2 mutations were observed only in patients>18 years. Patients over 65 years exhibited exclusively IDH1 R132H mutation. The proportion of oligodendroglioma patients increased with age (r=0.199, P=0.001). Tumor location was associated with IDH mutation type and tumor type (r=0.118, -0.077,P<0.05), but not with WHO grade. Compared to supratentorial tumors, non-R132H IDH1 mutations were significantly more common in infratentorial tumors (χ2=74.285, P=0.001), while IDH2 mutations rarely occurred infratentorially. Oligodendrogliomas were rarely found in infratentorial locations. IDH mutation type was associated with tumor type (r=0.118, P=0.001), but not with WHO grade. The incidence of non-R132H IDH1 mutations was significantly higher in astrocytoma than oligodendroglioma. The proportion of IDH2 mutations was significantly higher in oligodendrogliomas than astrocytoma (χ2=50.661, P<0.001). Conclusions: IDH mutations predominantly occur in adults, with an average age range of 25-55 years, although there are also rare pediatric cases. IDH-mutant gliomas are most located in the supratentorial region, particularly in the frontal lobe, and predominantly with IDH1 R132H mutation. Patients aged >65 years exclusively exhibit IDH1 R132H mutations. Compared to supratentorial tumors, non-R132H IDH1 mutations are significantly more frequent in infratentorial tumors. IDH2-mutant gliomas almost exclusively occur in adults and in supratentorial locations, with a significantly higher proportion in oligodendrogliomas than astrocytoma.

Genome-wide DNA methylation profiling, a high-throughput epigenomic technology, systematically examines the methylation status of CpG sites across the genome to generate "methylation fingerprints" that reflect cellular origins and developmental states. This technique has emerged as a transformative tool in the diagnosis of central nervous system (CNS) tumors. This article introduces the principles of this technology, along with its significant applications and major limitations in the diagnosis of CNS tumors. It proposes that current pathological diagnosis of CNS tumors should adopt an integrated diagnostic model centered on the pathologist, emphasizing that methylation profiling serves as an adjunctive tool rather than a replacement. Any methylation-derived result must be interpreted within the comprehensive context of clinical, radiological, histopathological, and molecular data. In the future, with advances in explainable artificial intelligence, multi-omics integration, and international standardization efforts, this technology is expected to evolve from a "revolutionary tool" into an "intelligent assistant" deeply integrated with pathological diagnosis.

Although homoscedasticity is often assumed in linear regression, real data may show variance patterns or residual structures that violate this assumption. We propose VarGuid, a variance-guided framework for two related settings: Covariate-dependent conditional variance under a global linear mean model, and residual nonlinear mean structure that can mimic heteroscedasticity. The framework has two deliberately separated components. The first uses an iteratively reweighted regression (IRR) algorithm to estimate a sparse global linear mean-variance model and support coefficient interpretation. The second uses a biconvex artificial-grouping algorithm for conditional prediction, keeping the fitted linear backbone fixed while adding group-specific local intercept corrections. We establish predictive-risk guarantees for the global estimator, and simulations and empirical studies show improved out-of-sample accuracy. VarGuid is illustrated in two applications: Health-related quality of life in low- and middle-income countries, and high-dimensional genomic prediction of lymph node evaluation in breast cancer.