Pancreas Tumour biology
ESMO Clinical Practice Guideline Express Update on daraxonrasib in the treatment of metastatic pancreatic cancer.
Resumen no disponible en la fuente.
Open oncology research index
Research evidence, connected and traceable.
Search scientific publications, inspect retrieval evidence and return to the original source without losing context.
- Records
- 81
- Research areas
- 7
- Sources
- 2
- Open access
- 41
Live index · Index updated
01 / Search
Search the evidence
Hybrid lexical and semantic retrieval across connected research sources.
81 results
Prostate Tumour biology
Prostate Cancer Care for Men with an Intellectual Disability: A Population-based Cohort Study of Symptoms, Diagnosis, Treatment, and Survival.
<h4>Background and objective</h4>Intellectual disability (ID) is increasingly recognised as a hidden driver of cancer mortality. However, evidence on prostate cancer (PC) care in this population is limited.<h4>Methods</h4>The study population comprised 29 554 men with an ID and 518 739 comparators from the Clinical Practice Research Datalink Aurum database, which is linked to hospital, mortality, and cancer registry data. Poisson and Cox regression analyses were used to estimate incidence rate ratios (IRRs), risk ratios (RRs), and hazard ratios (HRs) with 95% confidence intervals (CIs) for outcomes related to PC presentation, diagnosis, treatment, and survival.<h4>Key findings and limitations</h4>The ID group presented more frequently with symptoms suggestive of PC (IRR 1.35, 95% CI 1.28-1.43) but were less likely to have a prostate-specific antigen (PSA) test within 90 d (RR 0.66, 95% CI 0.63-0.70). Following detection of elevated PSA, the ID group had fewer referrals (RR 0.83, 95% CI 0.72-0.96), biopsies (RR 0.54, 95% CI 0.41-0.71), and PC diagnoses (RR 0.51, 95% CI 0.41-0.65). The ID group were also more likely to be diagnosed on the date of death (RR 5.96, 95% CI 2.70-11.77), have missing Gleason scores (RR 1.61, 95% CI 1.27-2.01), and present with de novo metastatic PC (RR 1.79, 95% CI 1.15-2.77). Among those with Gleason scores, the rate of clinically significant PC (Gleason ≥7) was comparable between the ID and control groups, while receipt of radical treatment for nonmetastatic PC was slightly lower in the ID group (RR 0.73, 95% CI 0.51-1.00). Men with an ID had twofold higher risk of death from PC following diagnosis (HR 2.11, 95% CI 1.64-2.73).<h4>Conclusions and clinical implications</h4>Men with an ID face disparities across the PC care pathway from investigation of relevant symptoms to survival after diagnosis. Targeted interventions are needed to address these inequities.
Pan-cancer Tumour biology
Topic-Driven Bibliometrics and Trend Intelligence for Stem Cell and Cancer Research.
The rapid growth of biomedical literature has created an urgent need for computational tools that enable researchers to systematically analyze publication trends, identify emerging research themes, and map the evolution of scientific fields. PubMed Atlas is a command-line and web-enabled workflow for topic-driven bibliometrics and trend intelligence using PubMed E-utilities. The pipeline executes PubMed queries, retrieves matching PMIDs, downloads full metadata records in batches, parses structured information (title, abstract, authors/affiliations, MeSH terms, publication types, grants, keywords, DOI), and stores normalized data in a local SQLite database for rapid querying and visualization. A Streamlit dashboard provides interactive exploration of publication trends, journal distributions, MeSH term summaries, geographic distributions, and recent article browsing with direct PubMed links. This protocol describes the installation, configuration, and operation of PubMed Atlas for cancer stem cell and stem cell transcriptional network research, and other fields, enabling investigators to conduct reproducible bibliometric analyses and identify knowledge gaps in rapidly evolving fields.
Pan-cancer Precision oncology
Computational Network Analysis for Defining Transcriptional Programs.
Cancer cell identity is governed by coordinated transcriptional programs that are frequently rewired during tumorigenesis. Systematic identification of cancer type-specific gene regulatory networks provides a framework for understanding oncogenic state transitions and for prioritizing candidate therapeutic targets. Here, we present a reproducible network-based workflow for reconstructing and analyzing transcriptional regulatory programs across human cancer types using publicly available expression datasets. We describe procedures for curating and preprocessing microarray data from the Gene Expression Omnibus, implementing random forest classification, and reconstructing gene regulatory networks using the CellNet platform. Detailed guidance is provided for evaluating classifier performance, quantifying network influence scores, integrating transcription factor, target interaction resources, and performing functional enrichment analyses. In addition, we outline approaches for comparing cancer-specific networks with corresponding normal tissue profiles to identify candidate drivers of malignant cell identity and potential prognostic biomarkers. Together, these protocols provide investigators with a scalable computational framework for defining cancer type-specific transcriptional states and for systematically interrogating regulatory mechanisms underlying tumor heterogeneity.
Pan-cancer Tumour biology
Identification of Genome-Wide Chromatin Structural Aberration in Cancer by Hi-C Analysis.
Aberrant three-dimensional genome organization is a hallmark of cancer, often driving oncogene activation through mechanisms such as enhancer hijacking. High-throughput chromosome conformation capture (Hi-C) maps these interactions on a genome-wide scale. Unlike earlier dilution-based methods, in situ Hi-C performs proximity ligation within intact nuclei, minimizing random ligation noise and enabling fine-scale structure detection. This chapter describes an optimized in situ Hi-C protocol tailored for cancer cell lines using MboI digestion and biotin-mediated pull-down to generate high-complexity libraries. We further outline a computational workflow that extends beyond standard topological mapping of compartments and topologically associating domains to identify cancer-specific aberrations. Specifically, we focus on detecting chromosomal rearrangements (structural variants) and characterizing the distinct circular topology of extrachromosomal DNA. This integrated experimental and analytical framework provides the necessary tools to dissect the spatial dysregulation underlying tumor evolution.
Pan-cancer Tumour microenvironment
Systematic Analysis of Tumor Microenvironment Using IOBR.
The Immuno-Oncology Biological Research (IOBR) package is an R-based analysis tool for exploring the tumor microenvironment (TME) and its influence on anti-tumor immunity. Built for high-throughput data-spanning both transcriptomic and genomic profiles-IOBR integrates six analytical modules, including transcriptomic data preprocessing, TME profiling, TME pattern identification, ligand-receptor interaction analysis, genome-TME interaction assessment, and visualization. In this chapter, we walk through a multi-omics workflow using example datasets, illustrating data preparation, distribution analyses, result interpretation, and graphical output. IOBR is open source and is available at https://github.com/IOBR/IOBR and a detailed GitBook ( https://iobr.github.io/book/ ) offers a complete manual and analysis guide for each function.
Colorectal Early detection
Automated Quantification and Morphometric Analysis of Pluripotent Stem Cell Colonies Using ColonyQuant.
Alkaline phosphatase staining is routinely used to evaluate the undifferentiated state of embryonic stem cells, yet quantitative interpretation of colony assays frequently depends on manual inspection and categorical scoring that introduce subjectivity and limit scalability. ColonyQuant provides a standardized, automated framework for objective analysis of alkaline phosphatase-stained colonies from conventional bright-field images. The software performs adaptive colony detection, per-colony intensity measurement, and extraction of eight geometric descriptors that collectively characterize colony size, compactness, symmetry, and boundary complexity. Feature tables generated by the workflow are structured to support statistical comparison across replicates and experimental conditions. Integrated analysis modules enable dimensionality reduction, supervised classification, and feature ranking, facilitating interpretation of phenotypic differences without requiring custom scripting. Visualization routines generate distribution plots, contour-density maps, multivariate embeddings, and representative shape mosaics to summarize population heterogeneity and morphological organization. Applied to pluripotent stem cell cultures subjected to chromatin perturbation, the platform detects coordinated changes in colony growth behavior and structural architecture that may not be evident through visual scoring alone. The protocol describes installation, configuration, batch image processing, quality control, hierarchical data aggregation, and downstream statistical analysis, providing a reproducible approach for transforming qualitative colony assays into quantitative, high-content phenotypic datasets suitable for stem cell research and screening applications.
Colorectal Tumour biology
Alkaline Phosphatase Staining of Embryonic Stem Cells.
Alkaline phosphatase (AP) is a widely utilized histochemical marker for assessing the pluripotent state of embryonic stem (ES) cells. Mouse ES cells maintained in an undifferentiated state exhibit high levels of tissue-nonspecific alkaline phosphatase (TNAP) activity, which sharply declines upon differentiation. AP staining provides a rapid, cost-effective, and reliable method to evaluate ES cell colony morphology, confirm maintenance of pluripotency during routine culture, and monitor the efficiency of differentiation protocols. Here, we describe a detailed protocol for AP staining of mouse ES cells using a chromogenic azo dye coupling method with naphthol AS-BI phosphate substrate and Fast Red Violet LB diazonium salt. This technique produces a vibrant red-purple precipitate in undifferentiated ES cell colonies, while differentiated cells remain unstained, enabling clear visualization and documentation of pluripotency status. The protocol includes optimized fixation conditions that preserve AP enzymatic activity, streamlined staining procedures, and guidelines for image acquisition and interpretation.
Breast AI and pathology
Advanced Image Generation for Cancer and Stem Cell Biology Using Diffusion Models.
Deep learning has transformed medical image analysis, but progress in cancer and stem cell applications is often constrained by limited access to large, diverse, well-annotated imaging datasets. This bottleneck is especially acute for studies of tumor heterogeneity and cancer stem cell (CSC) biology, where rare phenotypes and dynamic cell-state transitions-frequently linked to stemness-associated transcriptional programs (e.g., OCT4, SOX2, NANOG)-benefit from high-quality imaging across many samples and conditions. At the same time, regulatory and practical barriers (patient privacy, acquisition cost, and uneven institutional data sharing) restrict dataset scale and reuse. Diffusion models offer a practical route to synthetic data expansion by generating high-fidelity synthetic images that retain salient radiologic and pathologic features. In this chapter, we present an end-to-end protocol for adapting latent diffusion (Stable Diffusion) to oncology imaging using DreamBooth fine-tuning with small numbers of representative images, coupled with text-to-image and image-to-image workflows to generate controlled variations across modalities and disease presentations (e.g., brain tumor MRI, breast cancer mammography/CESM). We also describe quantitative and qualitative evaluation strategies, including Fréchet Inception Distance (FID) benchmarking and expert review considerations, to assess realism and diversity. These methods enable cancer and stem cell biologists to augment training data for segmentation and classification, build shareable educational resources, and prototype analyses for rare tumors or stemness-enriched subtypes while potentially reducing reliance on direct sharing of patient images.
Pan-cancer Tumour biology
Lentiviral Transduction of Embryonic Stem Cells.
Lentiviral vectors provide an efficient and reliable method for stable gene knockdown in embryonic stem cells (ESCs) through RNA interference. Here, we describe a detailed protocol for lentiviral transduction of mouse ESCs using lentiviral shRNA expression vectors. The protocol encompasses lentiviral particle production in HEK-293T packaging cells, determination of viral titer, transduction of ESCs cultured under feeder-free conditions, and selection of stably transduced cells. Additionally, we describe methods for evaluating transduction efficiency using fluorescence microscopy and flow cytometry, as well as for assessing gene knockdown efficacy by quantitative real-time PCR (Q-RT-PCR). This protocol is suitable for functional genomic studies in pluripotent stem cells and can be adapted for other difficult-to-transfect cell types.
Pan-cancer Precision oncology
Teratoma Formation and Genomic Profiling Using Multi-Omics Approaches.
Teratoma formation is the gold standard assay for evaluating the developmental pluripotency of human and mouse embryonic stem cells (ESCs) and induced pluripotent stem cells (iPSCs). Following subcutaneous injection into immunodeficient mice, pluripotent stem cells spontaneously differentiate into derivatives representing all three embryonic germ layers-ectoderm, mesoderm, and endoderm. Beyond serving as a functional assay for pluripotency, teratomas provide a unique three-dimensional model system for studying early human development and lineage specification in vivo. This chapter describes comprehensive protocols for teratoma formation in immunodeficient mice, tissue processing for multiple downstream genomic applications, and multi-omics profiling approaches. We detail methods for embryonic stem cell culture, teratoma generation via subcutaneous injection, tissue dissection and processing for chromatin immunoprecipitation followed by sequencing (ChIP-Seq), RNA sequencing (RNA-Seq), single-cell multiome profiling combining chromatin accessibility (ATAC-Seq) and gene expression (scRNA-Seq), and histological analysis using hematoxylin and eosin (H&E) staining. Additionally, we provide bioinformatics workflows for analyzing the resulting genomic datasets to characterize the epigenetic and transcriptional landscapes of teratoma-derived tissues. These methods enable comprehensive molecular characterization of developmental processes and provide valuable resources for stem cell biologists studying pluripotency, differentiation, and early embryonic development.
Lung Immunotherapy
Thymic Radiation is Associated with Worse Outcomes in Patients with NSCLC.
<h4>Background</h4>Emerging evidence indicates the functional importance of the thymus in adult health and may influence oncologic outcomes, yet current radiotherapy (RT) practice does not consider the thymus as an organ of interest. Because RT may incidentally expose the thymus to radiation, we hypothesized that thymic radiation dose is negatively associated with major clinical outcomes.<h4>Patients and methods</h4>This multicohort analysis included 1,107 patients with non-small cell lung cancer (NSCLC), including the phase III RTOG-0617 trial (n=460) and two independent real-world cohorts of patients treated with chemoradiotherapy alone (n=422; HARVARD-CRT) or in combination with consolidation immunotherapy (n=225; HARVARD-DURVA). We quantified thymic function using a deep-learning system that assessed thymic radiographic characteristics as a proxy for thymic function. Mean Thymic Dose (MTD) was used to measure thymic radiation exposure.<h4>Results</h4>Incidental thymic irradiation was associated with increased risk of distant metastases after confounding adjustments. Concretely, a 1Gy increase in MTD was associated with an increased distant metastasis risk of 1.57-4.21% (RTOG-0617: adjusted hazard-ratio [aHR]=1.29; P=0.0028; HARVARD-CRT: aHR=1.33; P=0.011; HARVARD-DURVA: aHR=1.95; P=0.007). In Thymic-Health-stratified analysis, patients with preserved Thymic Health prior to radiotherapy appeared to be at particularly greater risk, whereas no significant associations emerged in patients with impaired Thymic Health. One-year follow-up imaging demonstrated dose-dependent declines in thymic health and lower circulating lymphocyte counts, consistent with a possible biological link between thymic irradiation and immune competence loss. An exploratory feasibility study suggested that re-optimizing RT planning for thymic sparing can be achieved without compromising tumor coverage or cardiopulmonary constraints.<h4>Conclusions</h4>Thymic radiation exposure was independently associated with higher risks of metastasis and death in NSCLC patients, especially in those with preserved thymic function. These findings raise awareness towards considering the thymus as an organ of interest in radiotherapy and suggest that thymus-sparing strategies may help preserve immune health and improve patient outcomes.
Pan-cancer Tumour biology
[THE INDICATORS OF MORBIDITY, MORTALITY, ONE-YEAR LETHALITY AND DESOLATION OF PATIENTS WITH CERVICAL CANCER IN THE OSH OBLAST OF THE KYRGYZ REPUBLIC].
The purpose of the study is to evaluate morbidity, mortality, one-year lethality, and desolation of patients with cervical cancer in the Osh Oblast of the Kyrgyz Republic. The data about patients with cervical cancer in 2010-2023 was included in the study. The average value of morbidity of cervical cancer in the Osh region of the Kyrgyz Republic (over 14 years) made up to 10,3 ±3,6%000. The growth rate of indicator of morbidity of cervical cancer was 51,2%. The increase in mortality from cervical cancer was unsatisfactory in terms of the onco-gynecological service from 0.3%000 in 2010 to 5.3%000 in 2023. The growth rate was very high, exceeding 1 600%. The improvement and refinement of comprehensive therapy of patients with cervical cancer was evidenced by decrease of one-year mortality from 41.0% to 13.0%. The rate of decline was 68.3%. The positive trend is also observed in the rate of desolation of patients with cervical cancer, which decreased from 62.0% to 24.0%, as result of increased number of patients with early stages of the disease in the Osh region. The rate of desolation of patients with cervical cancer was 61.3%. The results of the study can be used to formulate proposals improving the organization of medical care support for patients with cervical cancer, as well as for formation of regional projects or departmental target programs.
Breast Precision oncology
<b>PGR expression as a pharmacogenomic companion biomarker to GENE70-derived genomic risk in ER-positive/HER2-negative breast cancer</b>.
<h4>Background</h4>The biology of the estrogen receptor-positive (ER+) and human epidermal growth factor receptor 2-negative (HER2-) breast cancers is heterogeneous even when they are categorized by their risk via genomics. Transcriptomic PGR expression reflects endocrine pathway activity and may provide complementary biological information within established GENE70-derived genomic-risk categories. Whether this molecular marker improves the biological interpretation of genomic-risk stratification beyond conventional clinicopathological assessment remains uncertain.<h4>Objectives</h4>The aim of this study was to determine whether transcriptomic PGR expression provides complementary biological and prognostic information within reconstructed GENE70-derived genomic-risk categories and refines the characterization of endocrine-related tumour biology in ER-positive/HER2-negative breast cancer.<h4>Methods</h4>This study analysed publicly available transcriptomic and clinical data from three cohorts: METABRIC (discovery cohort), GSE96058/SCAN-B cohort (validation cohort) and TCGA-BRCA cohort (molecular validation cohort). The GENE70-derived genomic-risk score was reconstructed for each cohort using matched genes. Cox regression, Kaplan-Meier analysis and subgroup comparisons were used to assess relationships between PGR expression, clinicopathologic variables, molecular features and survival outcomes.<h4>Results</h4>Across the three independent cohorts, low transcriptomic PGR expression was consistently associated with higher GENE70-derived genomic risk, increased MKI67 expression, reduced ESR1 expression and enrichment of the Luminal B subtype. Survival findings differed between cohorts. In the discovery METABRIC cohort, transcriptomic PGR expression showed heterogeneous associations with survival, particularly within GENE70-derived high-risk subgroups, whereas the external GSE96058/SCAN-B validation cohort demonstrated consistent associations between low PGR expression and poorer overall survival in both the overall ER-positive/HER2-negative population and GENE70-derived high-risk subgroups.<h4>Conclusion</h4>These findings suggest that transcriptomic PGR provides complementary biological and prognostic information within GENE70-derived genomic-risk categories. However, because treatment response was not evaluated in the present study, the findings should not be interpreted as evidence of predictive or pharmacogenomic utility and prospective studies incorporating treatment-response analyses are required before such applications can be established.
Pan-cancer Precision oncology
<b>Development of a biomarker-guided chemotherapy response model using combined CEA and CA19-9 profiles in esophageal cancer: A real-world pharmacotherapeutic study</b>.
<h4>Background</h4>Esophageal adenocarcinoma is associated with poor survival despite advances in systemic therapy. Readily available serum biomarkers may improve risk stratification; however, their combined role in biomarker-guided treatment decisions remains insufficiently investigated.<h4>Objectives</h4>To evaluate the prognostic significance of combined baseline carcinoembryonic antigen (CEA) and carbohydrate antigen 19-9 (CA19-9) levels and explore their utility for biomarker-based risk stratification in patients with esophageal adenocarcinoma receiving systemic chemotherapy.<h4>Methods</h4>This retrospective real-world cohort study included 38 patients with histologically confirmed esophageal adenocarcinoma who received systemic chemotherapy at a tertiary referral center between January 2016 and December 2021. Patients were classified into three exploratory biomarker-defined risk groups according to baseline serum CEA and CA19-9 levels. Overall survival was evaluated using Kaplan-Meier analysis and Cox proportional hazards regression, while treatment outcomes were explored across biomarker-defined groups and chemotherapy regimens.<h4>Results</h4>Elevated baseline CEA and CA19-9 levels were associated with shorter overall survival. Patients with simultaneous elevation of both biomarkers demonstrated the poorest survival outcomes (multivariable hazard ratio = 2.22). Differences in survival were observed across chemotherapy regimen groups within biomarker-defined categories; however, these comparisons were exploratory and may have been influenced by treatment-selection bias and unmeasured clinical factors. Patients with distal or gastroesophageal junction adenocarcinoma showed more favorable outcomes than those with cervical or thoracic tuxzmors.<h4>Conclusion</h4>Combined assessment of baseline CEA and CA19-9 may provide a practical and inexpensive approach for biological risk stratification in esophageal adenocarcinoma. However, these findings do not establish that the biomarkers can guide chemotherapy selection. Prospective multicenter studies with independent validation are required to determine their prognostic and predictive value before biomarker-guided treatment selection can be recommended in routine clinical practice.
Lung Immunotherapy
<b>The influence of LINC00511 and MDM2/MDM4 in lung carcinoma: Implications for immunological therapeutic strategies</b>.
<h4>Background</h4>Immunotherapy has significantly improved the treatment of lung cancer, but some patients experience aggressive disease, the mechanisms of which remain unclear. In particular, the interaction between lncRNA and the MDM2/MDM4 pathway warrants in-depth exploration.<h4>Objectives</h4>This study aimed to investigate the role of LINC00511 in promoting aggressive disease during immunotherapy for lung carcinoma by modulating MDM2/MDM4 expression.<h4>Methods</h4>This study collected tissues from 30 lung cancer patients undergoing immunotherapy and used A549 cell lines. RT-PCR detected expression of LINC00511, while immunohistochemistry and immunofluorescence detected MDM2/MDM4 expression. Bioinformatics analysis and luciferase reporter assays were performed to confirm their relationship. Cell migration, invasion, proliferation and response to PD-1 inhibitors were assessed.<h4>Results</h4>using A549 cells and female BALB/c nude mice (n=6 per group). LINC00511 was significantly upregulated in aggressive disease tissues and positively correlated with tumor stage. MDM2/MDM4 expression was also elevated (P < 0.01). LINC00511 directly targeted MDM2/MDM4. Silencing LINC00511 effectively inhibited tumor cell proliferation, migration, invasion and Ki-67 expression in vitro and in vivo . In the xenograft model, BALB/c nude mice treated with Nivolumab (0.1 mg/kg i.p. every 3 days) showed reduced tumor volume (mean ± SD: 445.6 ± 45.3 mm3 vs. 720.4 ± 80.2 mm3 in controls; P < 0.01).<h4>Conclusion</h4>LINC00511 promotes hyperprogressive disease in lung cancer during immunotherapy by upregulating MDM2/MDM4. Inhibition of LINC00511 effectively reverses tumor progression, suggesting a potential therapeutic target.
Breast Tumour biology
Design, Synthesis, In Vitro Biological Evaluation, and In Silico Studies of Novel Imidazo[4,5-b]Pyridine-Acrylonitrile-Based Derivatives as VEGFR-2 Inhibitors Against Breast Cancer and Hepatocellular Carcinoma.
Vascular endothelial growth factor receptor-2 (VEGFR-2) is a vital mediator of angiogenesis. Therefore, VEGFR-2 inhibition is considered a promising therapeutic target to combat cancer. In the present study, a series of 22 imidazo[4,5-b]pyridine-acrylonitrile-based derivatives was designed and synthesized. All compounds were evaluated for their VEGFR-2 inhibitory activity. Seven of the tested compounds showed high inhibitory activity against VEGFR-2 with IC<sub>50</sub> (0.029-0.087 µM) compared to reference drug sorafenib IC<sub>50</sub> = 0.091 µM. Four of these derivatives were selected for in vitro cytotoxic activity against breast cancer (MCF-7) and hepatocellular (HepG2) carcinoma cell lines, showing IC<sub>50</sub> (0.073-0.196 µM) and (0.20-0.21 µM), respectively, relative to sorafenib IC<sub>50</sub> 0.16 and 0.19 µM, respectively. Compound 2f exhibiting a superior VEGFR-2 inhibition (IC<sub>50</sub> = 0.029 µM compared to sorafenib IC<sub>50</sub> = 0.091 µM) and cytotoxicity against MCF-7 and HepG2 (IC<sub>50</sub> = 0.073 µM and IC<sub>50</sub> = 0.09 µM, respectively), was subjected to cell cycle analysis, apoptosis assay, and molecular modeling studies, which strongly supported the results. It caused cell cycle arrest at G2/M phase by 13.67% and a promising apoptosis induction. 2f exhibited a promising binding score and pose inside the VEGFR-2 active sites. It also showed a good safety profile with a moderate selectivity index. Based on the prediction of physicochemical and pharmacokinetic properties for 2f, it showed excellent prediction results to be orally bioavailable.
Pan-cancer Tumour biology
Gypenosides Damulin A and Damulin B Inhibit Hepatocellular Carcinoma by Regulating Cholesterol Synthesis.
Hepatocellular carcinoma (HCC) is a common malignant tumor of the digestive system. The liver is the primary organ for cholesterol production and metabolism in the body. Abnormal cholesterol levels can promote the initiation and progression of liver cancer, as well as influence treatment and patient prognosis. Damulin A and damulin B, a pair of isomeric dammarane-type saponins isolated from heat-treated Gynostemma pentaphyllum, have been shown to inhibit HCC cell proliferation and migration, arrest the cell cycle at the G0/G1 phase, induce apoptosis, and reduce intracellular cholesterol levels in vitro. Damulin B exhibited more potent effects in these assays. RNA sequencing and Western blot analysis revealed that both compounds downregulate the expression of key cholesterol biosynthesis-related genes, namely isopentenyl-diphosphate delta isomerase 1 (IDI1) and geranylgeranyl diphosphate synthase 1 (GGPS1). However, damulin A uniquely increased the expression of other cholesterol pathway genes: farnesyl diphosphate synthase (FDPS), farnesyl-diphosphate farnesyltransferase 1 (FDFT1), and lanosterol synthase (LSS). These findings indicate that damulin A and damulin B regulate intracellular cholesterol biosynthesis through distinct mechanisms, which may account for their differential inhibitory effects on hepatocellular carcinoma.
Pan-cancer Tumour biology
Robot-Assisted Radical Cystectomy: Recommendations from Latin American Experts Consensus on Surgical Strategy, Urinary Reconstruction, and Oncological and Func-tional Outcomes.
<h4>Purpose</h4>To develop contemporary evidence-informed recommendations for robot-assisted radical cystectomy (RARC), urinary reconstruction, and perioperative management, integrating current evidence with the experience of Latin American experts in uro-oncology and robotic surgery.<h4>Materials and methods</h4>A modified Delphi consensus process was conducted involving 42 experts with extensive experience in RARC. Topics included patient selection, perioperative optimization, enhanced recovery after surgery (ERAS), lymph node dissection, urinary reconstruction, functional preservation, perioperative systemic therapy, complex clinical scenarios, and emerging robotic technologies. A comprehensive literature review was performed using Medline, Scopus, and Web of Science through November 2025, following PRISMA principles and incorporating recommendations from EAU, AUA/ASCO/SUO, and NCCN guidelines. Consensus was defined as ≥75% agreement.<h4>Results</h4>Consensus supported routine implementation of ERAS protocols, structured frailty and nutritional assessment, extended pelvic lymph node dissection, and perioperative systemic therapy in eligible patients. Intracorporeal urinary diversion was associated with improved recovery and lower wound-related morbidity. Nerve-sparing and organ-preserving approaches were recommended in selected patients to optimize continence, sexual function, and quality of life. Orthotopic neobladder reconstruction should be individualized according to oncologic safety, functional status, renal function, and patient preference. RARC was considered feasible in complex settings, including obesity, bulky lymphadenopathy, locally advanced disease, and salvage surgery, when performed in experienced high-volume centers. Areas without consensus included urinary drainage strategies and antibiotic prophylaxis duration.<h4>Conclusions</h4>This expert consensus provides recommendations for RARC and urinary reconstruction, aiming to standardize practice, optimize perioperative care, and improve oncological and functional outcomes. Future prospective studies are needed.
Breast Tumour biology
Polygenic Risk Scores: Towards Clinical Application.
<h4>Background</h4>Common diseases are largely caused by the combined contributions of genetic and environmental risk factors. On the genetic side, numerous variants contribute to disease susceptibility, resulting in a substantial polygenic component of disease risk. Many such variants have already been identified. Polygenic risk scores (PRS), which aggregate the effects of these variants, enable more precise estimation of an individual's disease risk. We review the current state of bench-to-bedside translation of PRS.<h4>Methods</h4>This review is based on publications retrieved by a selective PubMed literature search using predefined search terms relevant to PRS.<h4>Results</h4>Studies across a range of diseases have shown that PRS can markedly influence the assessment of disease risk, enabling the identification of individuals at high risk. The predictive value of PRS can be improved by integrating them into established clinical risk models; this is the area of application in which the implementation of PRS in clinical practice is currently most advanced. For example, studies on coronary heart disease (CHD) and on breast and pancreatic cancer have shown that integrating PRS into clinical risk models changes personalized recommendations for primary and secondary prevention. Depending on the study, this affects 24% (CHD) to 25% (breast cancer) of individuals in the respective clinical risk groups. A major limitation is that most PRS are based on data from European populations, and their predictive value may be lower in individuals of other ancestral backgrounds. Moreover, quality standards must be established to ensure that the potential of PRS can be effectively translated into routine clinical practice.<h4>Conclusion</h4>As genetic data continue to grow, the predictive value of PRS is expected to increase further. PRS are therefore likely to become an integral component of risk prediction in preventive and precision medicine.
Prostate Tumour biology
A Thermo-Fluid-Solid Model Thermal Damage Prediction in Laser Ablation Therapy.
<h4>Background</h4>Precise temperature control is critical for safe, effective laser ablation in minimally invasive tumour treatment, but current models lack full understanding of multi-physics coupling during the process.<h4>Methods</h4>Using poroelastic theory, we developed a fully coupled optical-thermal-fluid-solid model for high-fidelity temperature prediction to optimise surgical protocols. Ex vivo experiments in pork tissue with precise needle positioning and multi-point temperature measurements validated the model.<h4>Results</h4>Simulations showed thermal expansion in elastic tissues induces fluid backflow to the heating zone (thermal focussing). A smaller beam waist enhances optical penetration, while higher porosity reduces peak temperature and promotes diffusion. Validation yielded a maximum prediction error ≤ 1.73°C.<h4>Conclusions</h4>This study provides a theoretical basis for precise temperature control and tissue damage prediction in laser therapy, with particular relevance to prostate tumour treatment.
Pancreas Tumour biology
Correction: Identifying health conditions associated with an increased risk of pancreatic ductal adenocarcinoma at medium term in nationwide electronic health records of primary care physicians.
Resumen no disponible en la fuente.
Pan-cancer Tumour biology
Weird, Witty and Worthwhile: Exploring the Benefits of Mnemonic-Based Medical Education.
<h4>Background</h4>Mnemonics, both self-created and well established, are commonly used by medical students to boost their recall of knowledge. Previous studies have evaluated mnemonics' utility in medical education, but few have demonstrated positive effects on students' knowledge and confidence. Mnemonics are also more commonly shared directly between peers, as opposed to being formally taught. This study explores the impact of mnemonic-based, near-peer-led teaching on medical students' knowledge and confidence. It aimed to determine whether mnemonics enhance students' ability to answer questions and boost self-perceived confidence and future mnemonic use.<h4>Methods</h4>Over 8 months, 23 peer-led teaching sessions on commonly tested clinical medical topics were held for third-year medical students, with mnemonics incorporated into all sessions. Presession and postsession SBA questions and five-point Likert scales were used to assess change in students' knowledge and confidence, respectively. Postsession, Likert scales were used to assess students' self-rated utility of the mnemonics taught and their likely future use of mnemonics.<h4>Results</h4>Students demonstrated a significant improvement in SBA scores, particularly for mnemonic-linked questions (p < 0.05). Confidence levels significantly increased postteaching (p < 0.01). Students reported a higher likelihood of future mnemonic use, and a strong correlation was found between improved confidence and increased mnemonic use (p < 0.001).<h4>Conclusion</h4>Peer-led teaching with mnemonic integration significantly improved both knowledge, student confidence and likelihood of future mnemonic use. This study highlights the potential of both mnemonic-based and peer-led learning strategies in medical education.
Colorectal Early detection
Malassezia May Contribute to Basal Cell Carcinoma Development via Inflammatory and Oxidative Stress Pathways.
Basal cell carcinoma (BCC) is the most common malignant skin tumor. Skin-resident lipophilic Malassezia yeasts are associated with various cutaneous disorders, while their correlative patterns and potential biological effects in BCC tissues remain insufficiently defined. We used RT-qPCR screening of archived FFPE BCC specimens and metagenomic sequencing of three paired fresh tumor and peritumoral tissues to characterize tissue-associated Malassezia colonization. M. globosa was the most abundant species in FFPE samples and was also detectable in fresh tissues. In vitro functional assays (CCK-8, EdU) in HaCaT keratinocytes and A-431 epidermoid carcinoma cells showed that 12 h stimulation with optimal concentrations of M. globosa (1.2 × 10<sup>7</sup> CFU/mL) and M. yamatoensis (1.6 × 10<sup>7</sup> CFU/mL) significantly promoted epithelial cell proliferation. Transcriptome sequencing and subsequent RT-qPCR validation further showed that both strains significantly upregulate pro-inflammatory genes (IL-1β, IL-6, TNF-α) and oxidative stress-related genes (SOD1, SOD2) in these cell lines. Collectively, our findings describe a correlative association between Malassezia colonization and BCC lesions and offer preliminary in vitro mechanistic clues.
Lung Tumour biology
Primary pulmonary salivary gland-type tumors in cytopathology practice: A systematic review and meta-analysis.
<h4>Background</h4>Primary pulmonary salivary gland-type tumors (PSGTs) are rare but clinically significant tumors that originate from the submucosal glands of the tracheobronchial tree. Cytologic samples taken during bronchoscopy are a key component of preoperative evaluation. However, cytologic diagnosis remains challenging because of the submucosal growth and morphologic overlap of PSGTs. In addition, current knowledge of the cytohistologic correlation of PSGTs is fragmented. The objective of this study was to assess the effectiveness of cytologic diagnoses of PSGTs.<h4>Methods</h4>A comprehensive, systematic literature search of the PubMed database was conducted to identify studies with cytologic and histologic diagnoses of PSGTs. Comprehensive data on diagnostic and clinical factors, when available, were collected for all individual patients. The data were tabulated in Microsoft Excel and analyzed using OpenMeta (Analyst) software.<h4>Results</h4>In total, 49 studies comprising 106 patients were identified. Final cytohistologic concordance was demonstrated in 48.1% of cases. Fine-needle aspiration showed the highest sensitivity (75.0%), followed by bronchial/tracheal washing (38.1%), and bronchial brushing (34.2%). Adenoid cystic carcinoma was the most common histologic subtype, accounting for 67 cases, followed by mucoepidermoid carcinoma, which accounted for 27 cases.<h4>Conclusions</h4>The cytologic diagnosis of rare PSGTs remains challenging. Overall, cytohistologic concordance was 48.1%. However, fine-needle aspiration demonstrated greater diagnostic accuracy than exfoliative cytology and may facilitate a more accurate preoperative assessment.
Leukemia Tumour microenvironment
Outcomes of adult patients with Li-Fraumeni syndrome and myeloid neoplasms.
<h4>Background</h4>Li-Fraumeni syndrome (LFS) is an inherited cancer predisposition syndrome. Hematologic malignancies are not considered LFS defining tumors, however, both acute lymphoblastic leukemia and therapy-related myeloid neoplasms (MNs) in LFS are described. Treatment approaches and outcomes of MN in LFS need further evaluation.<h4>Methods</h4>The authors performed a retrospective analysis to understand treatment approaches and outcomes in patients with LFS who developed an MN.<h4>Results</h4>Among 190 patients with LFS between February 2001 and February 2026 with a history of at least one neoplasm, 14 (7%) had an MN. Median age at MN diagnosis was 43 years (range, 25-73) and 11 (79%) patients were female. Overall, eight (57%) patients had myelodysplastic syndrome (MDS), five (36%) had acute myeloid leukemia (AML), and one (7%) had T-myeloid mixed phenotype acute leukemia (MPAL). With frontline therapy, six (75%) patients with MDS and one (17%) patient with AML achieved an overall response. Considering all lines of therapy received, cumulatively six (75%) patients with MDS and five (83%) patients with AML/MPAL achieved an overall response. At a median follow-up of 28.8 months, the median overall survival (OS) was 18.2 months, and 1-year and 2-year OS rates were 77% and 17%, respectively. Five patients (two MDS, two AML, and one MPAL) underwent hematopoietic stem cell transplantation during their MN therapy with a median OS of 19.3 months.<h4>Conclusion</h4>Although short-lived responses to leukemia-directed therapy are common, long-term survival in most patients with LFS developing MN are poor. Additional research and understanding of the mechanisms to prevent MN and to improve MN treatment in LFS are needed.
Pan-cancer Resistance
The Crosstalk Between Diabetes and Alzheimer's Disease: A Molecular Perspective.
The global prevalence of type 2 diabetes mellitus (T2DM) and Alzheimer's disease (AD) is increasing significantly in an age-dependent manner. Growing evidence supports the conceptualization of AD as 'type 3 diabetes,' a term proposed to describe a metabolic disease primarily driven by impaired insulin signalling and insulin resistance within the brain. This review explores the shared molecular mechanisms underlying both T2DM and AD, including chronic neuroinflammation, oxidative stress, mitochondrial dysfunction and impaired glucose metabolism. Specifically, the crosstalk involves the phosphoinositide 3-kinase (PI3K)/protein kinase B (Akt) pathway, where insulin resistance leads to increased glycogen synthase kinase 3β (GSK-3β) activity, promoting tau hyperphosphorylation and amyloid-β (Aβ) accumulation. Furthermore, the article examines the roles of the NOD-like receptor protein 3 (NLRP3) inflammasome, O-linked β-N-acetylglucosamine modification (O-GlcNAcylation), and the gut-brain axis as critical mediators linking metabolic dysfunction to neurodegeneration. Unlike previous reviews that predominantly address individual pathways in isolation, this review provides an integrated molecular framework that connects insulin resistance to neurodegeneration through converging signalling cascades and highlights emerging therapeutic targets including the NLRP3 inflammasome and O-GlcNAcylation as potential mechanistic bridges between T2DM and AD. Finally, the therapeutic potential of various antidiabetic agents such as glucagon-like peptide-1 (GLP-1) receptor agonists, sodium-glucose cotransporter-2 (SGLT-2) inhibitors and thiazolidinediones is discussed, as these drugs offer promising opportunities for the treatment and prevention of AD by targeting these common molecular pathways.
Pancreas Tumour biology
AS160 Overexpression Protects Against β-Cell Failure in Hyperglycemic Mice by Interacting With HSPA8 to Regulate Mitophagy.
<h4>Aim</h4>Insulin deficiency due to pancreatic β-cell loss and dysfunction is a key event in the pathogenesis of T1DM and a progressive driver of T2DM. This study aims to investigate the role of AS160 in regulating mitochondrial homeostasis and insulin secretion in pancreatic β-cells, and to elucidate the underlying molecular mechanism involving its interaction with HSPA8 and activation of PINK1-Parkin-mediated mitophagy.<h4>Methods</h4>Insulin and AS160 levels in islets were analyzed by immunofluorescence staining. β cell-specific AS160 overexpression mice were generated via lentivirus injection, and their metabolic phenotypes were characterized. In vitro, AS160 was overexpressed or knocked down to assess its impact on cell proliferation, insulin secretion, and mitochondrial function. AS160-interacting proteins were identified by immunoprecipitation-mass spectrometry (IP-MS).<h4>Results</h4>AS160 expression was significantly decreased in islet and correlated positively with insulin levels in hyperglycemic mice. Specific overexpression of AS160 in β-cells exhibited novel protective effects for the islets and insulin levels in hyperglycemic mice. Mechanistically, AS160 overexpression in β-cells increased mitophagy and preserved mitochondrial biogenesis to maintain healthy mitochondrial homeostasis and insulin secretion. At the molecular level, HSPA8 was identified as a novel AS160-interacting protein that enhances PINK-dependent mitophagy. Knockdown of HSPA8 reversed the overexpression of AS160-induced mitophagy and mitochondrial biogenesis.<h4>Conclusion</h4>Collectively, this work identifies the AS160-HSPA8 interaction as a key mechanism that sustains mitochondrial homeostasis through regulating mitophagy and mitochondrial biogenesis, thus preserving β-cell mass and function. These findings suggest that AS160 emerges as a pivotal regulator of mitochondrial homeostasis in pancreatic β-cell in vivo.
Colorectal Tumour biology
Spontaneous Hepatobiliary Disease in a Breeding Colony of Aotus spp.
<h4>Background</h4>Owl monkeys are nonhuman primate species in the genus Aotus. The Michale E. Keeling Center for Comparative Medicine and Research maintains an Owl Monkey Breeding and Research Resource composed of four species of owl monkeys: Aotus nancymai, A. vociferans, A. azarae, and A. griseimembra. Disease of the hepatobiliary system has not been well characterized in Aotus.<h4>Methods</h4>Necropsy records from a 17-year period were reviewed to characterize the gross and microscopic lesions of liver disease.<h4>Results</h4>316 adult owl monkeys were submitted for necropsy in the review period; 22 of 316 (6.96%) were diagnosed with hepatobiliary disease. Cardiac-associated hepatopathy was the most common diagnosis (6/22). Previously unreported conditions included cystic mucinous hyperplasia (1/22), acute cholangiohepatitis (1/22), idiopathic end-stage hepatopathy (1/22), cholangiocarcinoma (2/22), and acute cholecystitis (4/22). Vitamin E deficiency anemia (3/22) and hemosiderosis (4/22) have been reported in the genus, and these diagnoses were noted.<h4>Conclusions</h4>The spectrum of hepatobiliary pathology in owl monkeys is broader than previous reports have suggested.
Breast AI and pathology
A Machine Learning Approach That Measures Extracellular pH Within an In Vivo Tumor Model Using acidoCEST MRI.
Quantitative mapping of extracellular tumor pH (pHe) using acidoCEST MRI offers a method to characterize the tumor microenvironment. Conventional analysis of acidoCEST MRI relies on fitting the Bloch-McConnell equations to experimental CEST spectra. However, this "Bloch fitting" method is computationally intensive. Recent work has demonstrated that machine learning can accurately predict pH from CEST spectra of phantoms containing iopamidol, providing an alternative to Bloch fitting for acidoCEST MRI analysis. In this study, we evaluated the ability of machine learning models to estimate extracellular tumor pHe using acidoCEST MRI in a 4T1 murine breast cancer model. Thirty-eight tumor-bearing mice were imaged using a CEST-FISP acquisition, and CEST spectra were extracted along with T<sub>1</sub>, T<sub>2</sub>, B<sub>1</sub>, and B<sub>0</sub> maps. Pixel-wise pH values derived from Bloch fitting served as reference values. We trained and tested three machine learning models: (1) random forest regression (RFR) using CEST spectra and T<sub>1</sub>, T<sub>2</sub>, B<sub>1</sub>, and B<sub>0</sub> maps; (2) RFR using only CEST spectra; and (3) a convolutional neural network (CNN) using only CEST spectra. The RFR model using only CEST spectra achieved the best performance, with mean absolute percentage errors of 0.29% (training) and 0.79% (testing), corresponding to 0.020 and 0.054 pH units, respectively. Spatial pHe maps generated by the RFR methods closely matched those from Bloch-McConnell fitting, whereas CNN-derived maps showed compressed pH ranges and reduced correlation with reference pH values. These findings demonstrate that RFR provides a computationally efficient alternative to Bloch fitting for in vivo acidoCEST MRI.
Lymphoma Tumour biology
Elevated Expression of Caveolin-1 Reflects a Poor Prognosis in Patients With Primary Central Nervous System Lymphomas.
Caveolin-1 (Cav-1) is a component of the vesicle-like structure of the plasma membrane called caveolae. Cav-1 is involved in the transport of substances from the plasma membrane to the intracellular space of various cells. It has been reported that it is involved in the growth mechanisms, angiogenesis, and immune evasion of various malignant tumors. In the present study, the relationships between Cav-1 expression in primary central nervous system lymphomas (PCNSLs) and clinicopathological characteristics were investigated to clarify the role of Cav-1 in the progression of PCNSLs and patient prognosis. The expression of Cav-1 was assessed in 39 PCNSLs (CNS diffuse large B-cell lymphomas) using immunostaining, and its prognostic significance was evaluated. Cav-1 expression by PCNSLs was evaluated using the histoscore (H-score). The H-score for Cav-1 in PCNSLs ranged from 0 to 90 (median 20); there were 12 cases showing no score (H-score 0), 11 cases with a low score (0 < H-score < 20), and 16 cases with a high score (H-score ≥ 20). Survival was significantly lower in patients with high H-scores than in those with no score and low H-scores (p < 0.01), and in patients with high numbers of Cav-1-positive vessels than in those with low numbers of Cav-1-positive vessels (p < 0.05). Cav-1 positivity in tumor cells was correlated with the Ki-67 labeling index and the number of Cav-1-positive vessels. These results suggest that Cav-1 may be involved in the proliferation and angiogenesis of PCNSL cells. The present study also demonstrated that overexpression of Cav-1 in PCNSLs is associated with tumor progression and a poor patient prognosis.
Pan-cancer Tumour biology
Relationship Between Hypovitaminosis C, Obesity, and Features of Metabolic Syndrome-A Narrative Review.
This narrative review aimed to review plausible mechanisms for the role of vitamin C (ascorbic acid [AA]) in the maintenance of healthy weight and energy metabolism; examine the evidence for inadequate vitamin C (plasma AA <50 µmol/L), hypovitaminosis C (≤23 µmol/L), and vitamin C deficiency (≤11.4 µmol/L) in the disrupted homeostasis of obesity and metabolic syndrome; and ascertain whether vitamin C supplementation or dietary intervention could potentially treat obesity and the associated features of metabolic syndrome. Vitamin C hypovitaminosis and deficiency are prevalent in developed countries, despite the widespread availability of vitamin C-containing fruit and vegetables and vitamin supplements. Western diets are characterized by highly processed, macronutrient-rich foods, which are deficient in dietary fiber and micronutrients. This contributes to postprandial oxidative stress and gut dysbiosis, leading to profound effects on insulin sensitivity, hyperglycemia, levels of endotoxemia, fatty acid oxidation, adipocyte hypertrophy, and regulation of metabolism and energy balance. The existing in vitro and in vivo preclinical data demonstrate the effectiveness of vitamin C as both a prophylactic and a therapeutic intervention for obesity and metabolic syndrome. The outcomes in human intervention studies are more modest, with improvements in insulin sensitivity, lipid profile, metabolic inflammation, weight, hypertension, gut permeability, and hepatic steatosis. Some clinical studies are limited by the lack of baseline plasma AA concentrations, or the inability to optimize plasma AA in participants with hypovitaminosis C or metabolic syndrome. The addition of vitamin C to physical activity and dietary interventions may improve the efficacy of treatments for obesity and metabolic dysfunction. However, more data are required to understand the synergism between vitamin C supplementation, medical nutrition therapy, adequate exercise, and pharmacological intervention in weight control and metabolic syndrome management.
Pan-cancer Early detection
Hyperpolarized MRI in Oncology-A Review of Clinical Studies and Evaluation of Preclinical Liver Studies.
Hyperpolarized MRI is a medical imaging technology that has been used so far only in focused clinical studies. This technology enables metabolic imaging without ionizing radiation and is combined with standard-of-care MRI protocols. Unlike metabolic imaging performed with radiolabeled materials (e.g., PET and SPECT), hyperpolarized MRI enables differentiation between the injected metabolic substrate and its product(s). In this way, the number of imaging contrasts that may be attained is relatively large. The purpose of this review is two-fold: to provide coherent terminology for the various contrasts obtained in HP-MRI to facilitate further development in medicine and to highlight its potential contribution to liver nodule characterization. We reviewed 33 hyperpolarized MRI papers published from 2013 to 2024 that describe clinical studies in human subjects and patients, and characterized the metabolic contrast provided by this technology across various organs and in cancer tumors. We have further reviewed 20 preclinical studies of hyperpolarized MRI in the liver to evaluate the potential of this technology for early detection and characterization of hepatocellular carcinoma. Most clinical studies have used a single hyperpolarized substrate, namely, pyruvate, labeled with carbon-13 at position 1 ([1-<sup>13</sup>C]pyruvate). Even with this single substrate, the number of possible imaging contrasts is high, making the interpretation of these images challenging. We aimed to explain these imaging contrasts and provide a consistent nomenclature for them that will serve the medical community. Preclinical MRI studies using hyperpolarized substrates suggest the potential to characterize liver nodules in patients. In summary, hyperpolarized MRI is a promising clinical imaging modality in oncology; however, the small number of patients investigated so far precludes addressing possible biological confounding factors. Further studies are required for improving tumor characterization using this technology, and studies in liver imaging aimed at early detection of HCC are warranted.
Pan-cancer Resistance
Sequence Inversion Dictates the Antiproliferative Activity of Bidirectional Tryptophan-Containing Dipeptide Libraries.
Two complementary series of novel tryptophan-derived dipeptides (Trp-X and X-Trp) were designed and synthesized to systematically evaluate how backbone sequence inversion alters chemical accessibility and in vitro antineoplastic profiles against a human solid tumor cell line panel (A549, HeLa, MIA PaCa-2, SW1573, T-47D, and WiDr). Reversing the peptide connectivity revealed prominent sequence-dependent chemical reactivities and biological variations. Steric repulsion at the β-carbon limited basic hydrolysis during X-Trp precursor assembly, whereas the Trp-X series allowed straightforward chemical couplings. Phenotypic screening demonstrated that the Trp-X configuration is biologically superior to the X-Trp layout. The conformationally restricted L-proline conjugate (Trp-Pro) emerged as the lead architecture, exhibiting consistent, single-digit sub-micromolar growth inhibition across all histotypes (GI<sub>50</sub> = 1.36-2.06 μM) and effectively outperforming clinical standards cisplatin and 5-fluorouracil in resistant models. Interestingly, an inversion of structure-activity relationships was observed in the reverse series, where specific residues like L-phenylglycine and L-tyrosine experienced a prominent rescue of potency upon sequence relocation. These results confirm that the antiproliferative profile of these peptidomimetics is strictly dictated by a highly directional, sequence-specific molecular topology rather than aggregate lipophilicity alone.
Breast Tumour biology
Expression of Concern: Regulation of the breast cancer stem cell phenotype by hypoxia-inducible factors.
Resumen no disponible en la fuente.
Pancreas AI and pathology
Pancreatic Cancer Screening and Early Detection: Where Are We Now?
Pancreatic cancer remains one of the most lethal malignancies, with an overall 5-year survival rate of approximately 13%. The poor prognosis is primarily due to late-stage diagnosis, as early-stage disease is often asymptomatic. Screening for pancreatic cancer in high-risk individuals offers the potential for early detection, timely intervention, and improved survival outcomes. Individuals at elevated risk for pancreatic cancer-that is, those with a family history or predisposing genes for pancreatic cancer or certain inherited syndromes and patients with precursor lesions or precancerous lesions-are considered suitable candidates for screening. Imaging is recognized as an important tool for diagnosis and staging of pancreatic cancer. Evidence supports use of endoscopic US, MRI, and CT as primary imaging modalities to facilitate early detection. Screening intervals and imaging strategies vary in high-risk individuals and those with precursor or precancerous lesions. While conventional imaging plays a key role in diagnosis and staging, it often fails to allow detection of visually occult disease, an area where artificial intelligence-enhanced imaging offers potential. These advancements have the potential to improve detection of early pancreatic cancer with prediagnostic CT and MRI. In parallel, the development of novel serum biomarkers with greater sensitivity and specificity provides promising avenues for noninvasive screening. Despite these advances, there remains a critical unmet need for validated, cost-effective, and scalable tools. The authors aim to critically appraise current screening strategies, highlight emerging technologies in imaging and biomarkers, and discuss their potential to transform early detection and improve clinical outcomes in high-risk populations. © The Author(s) 2026. Published by the Radiological Society of North America under a CC BY 4.0 license See the invited commentary by Luk and Flusberg in this issue.
Pancreas Early detection
Invited Commentary: Pancreatic Cancer Screening and Early Detection: Current Perspectives and Future Opportunity.
Resumen no disponible en la fuente.
Pediatric Tumour biology
Impact of post-chemotherapy classification on outcomes in stage III and IV unilateral favorable histology Wilms tumor: Pooled data from Children's Oncology Group AREN0532, AREN0533, and AREN03B2 studies.
<h4>Background</h4>Improvements in therapy for patients with favorable histology Wilms tumor (FHWT) have relied on refinement of risk stratification for therapeutic assignment through identification of favorable and unfavorable prognostic factors. Although well established in International Society of Pediatric Oncology-Renal Tumor Study Group protocols, the Children's Oncology Group (COG) has not historically used post-chemotherapy histology (PCH) to guide therapy. This study examines whether outcomes are associated with PCH classification in the COG treatment context.<h4>Methods</h4>The authors identified 2386 patients with stage III or IV FHWT enrolled on AREN0532, AREN0533, or AREN03B2-only from 2006 to 2019. Inclusion criteria included delayed nephrectomy more than 5 and fewer than 16 weeks after initial biopsy, treatment on or as per protocol with DD4A or Regimen M, and known PCH classification and outcome.<h4>Results</h4>Of 352 included patients, 45 (12.7%) were classified as low risk (LR), 287 (81.5%) as intermediate risk (IR), and 20 (5.6%) as high risk (HR). PCH classification was significantly prognostic for event-free survival (EFS) (p < .0001) and overall survival (OS) (p < .0001). Four-year EFS for LR, IR, and HR groups was 93% (95% CI, 86%-100%), 85% (95% CI, 81%-89%), and 53% (95% CI, 35%-81%); 4-year OS was 96% (95% CI, 90%-100%), 93% (95% CI, 89%-96%), and 62% (95% CI, 44%-89%). Compared with LR or IR, EFS and OS were significantly lower for HR, regardless of stage or treatment.<h4>Conclusion</h4>PCH is a significant prognostic factor in unilateral FHWT treated on COG regimens, strongly supporting its incorporation into therapeutic stratification in prospective COG trials.
Pan-cancer AI and pathology
Big data and psychiatry: advances, constraints and future directions.
Early work in psychiatry research, often involving single sites, small samples, and limited variables, has shifted to contemporary research involving multiple sites, large samples, and many variables. Such research raises important questions, including concerns about data quality and methodological rigor, uncertainty about its key lessons, issues regarding clinical relevance, and questions about how to optimize future advances. Here we consider these questions and concerns against the context of big data work on community and register-based surveys, cohort and biobank studies, electronic health records, digital phenotyping, brain imaging, genomics and other -omics, and randomized controlled trials. The development of large datasets allowing well-powered analyses is a major milestone, but sample size alone does not guarantee more precise estimates, and ongoing attention to the quality and rigor of big data collation and analysis is needed. Big data research has fostered trans-disciplinarity and given insights into mechanisms underlying psychiatric disorders, but also emphasizes the intricacy, heterogeneity and variability of such mechanisms, and the importance of triangulating between large-scale and small-scale research. The complexity of psychiatric phenotypes and psychobiological mechanisms contributes to the difficulty in bridging from big data to clinical application; big data research reinforces the importance of holding our diagnoses of psychiatric disorders lightly and providing explanations of these conditions humbly; and future work needs to be more attentive to clinical issues. There is enormous scope for further building databases relevant to psychiatry, but advances in conceptual models and asking the right questions are equally valuable. The full impact of big data, including artificial intelligence analyses, remains to be seen, but overenthusiastic support should be tempered by a better understanding of its strengths and limitations. At its best, such work will contribute in an iterative and integrative way to advancing our knowledge of psychiatric disorders and mental health.
Pan-cancer Tumour biology
Fertility-Sparing Treatment of Endometrial Cancer: Survey of Australian Practice.
<h4>Background</h4>Endometrial cancer is the most common gynaecological malignancy in Australia. Patients diagnosed in their reproductive window may desire a fertility-sparing approach, opposed to standard surgical care. International guidelines on fertility-sparing management exist; however, these highlight the lack of high-level evidence to support the recommendations. There are no Australian-specific guidelines or registries, leading to uncertainty as to how patients are managed locally.<h4>Aim</h4>To survey clinicians in Australia regarding fertility-sparing management of endometrial cancer, to identify areas of consensus and variation, barriers to care and research priorities.<h4>Materials and methods</h4>A 47-item online survey was distributed to clinicians who manage patients with endometrial cancer in Australia between November 2024 and January 2025. Responses were analysed descriptively.<h4>Results</h4>Of 70 eligible clinicians, 56% responded. There was agreement on appropriate candidate selection for FIGO grade 1 disease with absent/focal myoinvasion, use of hysteroscopy for diagnosis, and preference for levonorgestrel intra-uterine device as initial therapy. Opinions were divided on management of patients with Lynch syndrome or more extensive myoinvasion, the diagnostic utility of endometrial pipelle, and treatment of recurrences. POLE mutation testing and hysteroscopic resection were uncommon practices. Most respondents referred patients to fertility specialists, while other members of the multi-disciplinary team were less frequently utilised. The most frequent research priorities were weight-loss strategies (31%) and molecular profiling (27%).<h4>Conclusions</h4>Clinicians in Australia have common approaches to diagnosis and first-line therapy, but there is variation in patient selection, surveillance and engagement with a multi-disciplinary team. These discrepancies highlight the potential benefits of local guidelines and further research.
Pan-cancer Tumour biology
A Frequentist Approach Using Weighted Estimating Equations for Survival Analysis of Randomized Controlled Trial With External Control Arm.
Clinical trials using a hybrid control, which integrates a randomized control with external control data, may be one solution to facilitate clinical development where traditional randomized controlled trials are challenging. Despite its attractive features, systematic differences between controls can introduce bias in treatment effect estimation and inflate Type I error. Although propensity score methods may be used to address systematic differences from measured covariate imbalance, potential differences from unmeasured or unknown factors may still arise despite careful consideration. In such cases, dynamic borrowing methods have been proposed to account for potential differences between randomized and external control data and discount the contribution of external control data when needed. Conventional methods, however, often rely on Bayesian approaches, requiring proper specification of the likelihood. This is particularly challenging for survival analysis using proportional hazards model due to specification of the baseline hazard function. To address these challenges, we propose to utilize the weighted partial likelihood score equation where external control patients are weighted by the inverse probability of the propensity score and an additional discounting weight, adjusting the contribution of external control data according to its similarity to randomized control data. A new variance estimator based on stacked estimating equations is proposed to incorporate the uncertainty in estimating discounting weights. Operating characteristics of the proposed method are studied via extensive simulation. The proposed method yields reduced bias in treatment effect estimation and Type I error while improving statistical power in either scenarios where measured or unmeasured confounders exist.
Pan-cancer Tumour biology
Impact of 4D-CT MIP reconstruction method on iGTV definition: A clinical case report.
To present a clinical case demonstrating the impact of 4D-CT maximum intensity projection (MIP) reconstruction methodology on internal gross tumor volume (iGTV) definition and to highlight implications for motion management quality assurance. During routine thoracic radiation therapy treatment planning, a discrepancy in tumor extent was identified during physician contouring. The inferior extent of the tumor appeared artificially truncated on the MIP dataset used for target delineation. Further review revealed that the default MIP had been generated from phase-sorted 4D-CT images. A comparison was performed between the phase-sorted MIP and a MIP reconstructed from the original cine images to evaluate differences in motion representation. The phase-sorted MIP under-represented the full tumor motion envelope, most notably in the inferior direction. In contrast, the cine-based MIP demonstrated a more complete representation of tumor extent throughout respiration. This discrepancy was not readily apparent during the routine clinical workflow, and verification of the MIP reconstruction method was not part of the standard QA review process at the time. Differences in MIP reconstruction methodology can impact iGTV definition. Although these differences may be subtle in most cases, clinically meaningful discrepancies may occur, particularly in patients with irregular breathing patterns. This case highlights the importance of awareness and verification of 4D-CT MIP reconstruction methods and supports comprehensive QA across CT simulation, treatment planning, and motion management workflows.
Pancreas Early detection
Prevalence and Clinical Significance of Abnormal Findings on Pre-Diagnostic Imaging in Pancreatic Ductal Adenocarcinoma.
<h4>Objectives</h4>Pancreatic ductal adenocarcinoma (PDAC) is the third leading cause of cancer-related death. Although CT/MRI are used for initial evaluation, early imaging findings suggesting PDAC may be unappreciated, resulting in diagnostic delays. We aimed to determine the prevalence and types of pancreatic abnormalities on pre-diagnostic imaging in patients later diagnosed with PDAC.<h4>Methods</h4>We identified patients diagnosed with PDAC between 2022 and 2025 who had a pre-diagnostic abdominal CT/MRI performed 3-36 months before diagnostic imaging. These patients were compared with a non-PDAC cohort who underwent pancreas cancer screening but did not develop PDAC after 5 years of follow-up. The primary outcome was the prevalence and type of pancreatic abnormalities on pre-diagnostic scans.<h4>Results</h4>Among 733 PDAC patients, 151 (20.6%) had pre-diagnostic imaging (mean age 72.6 y, 53.6% F). At diagnosis, mean tumor size was 3.1 cm and 42.4% had metastatic disease. The mean (±SD) interval between pre-diagnostic and diagnostic scans was 15.9±9.5 months. Abnormal pancreaticobiliary findings were reported in 27.2% of pre-diagnostic scans. Pancreatic cyst prevalence was similar between pre-diagnostic and non-PDAC groups ( P =0.07). However, non-cyst pancreatic abnormalities were more frequent in PDAC pre-diagnostic scans ( P =0.04), including pancreatic duct dilation ( P =0.07) and pancreatic atrophy ( P =0.05).<h4>Conclusions</h4>Although >20% of PDAC patients had pre-diagnostic imaging, most pre-diagnostic scans reported no pancreatic abnormalities, despite eventual advanced disease at diagnosis. Subtle findings suggesting PDAC, such as duct dilation and pancreatic atrophy, were uncommon on pre-diagnostic scans but more frequent than in non-PDAC scans. These findings may represent early signs of PDAC and warrant closer surveillance.
Prostate Tumour biology
Editorial comment: FOCS-24: An integrated oncologic-functional endpoint and its preoperative nomogram after focal therapy for localized prostate cancer.
Resumen no disponible en la fuente.
Prostate Tumour biology
FOCS-24: An integrated oncologic-functional endpoint and its preoperative nomogram after focal therapy for localized prostate cancer.
Focal therapy (FT) seeks to balance prostate cancer (PCa) control with preservation of urinary and sexual function, yet most studies report these outcomes separately. We developed FOCS-24, a composite endpoint capturing full oncologic and functional recovery at 24 months after FT for PCa, and created a preoperative nomogram estimating its probability. From 2021 to 2025, 324 men underwent primary FT (cryoablation or HIFU) at a tertiary cancer center. FOCS-24 required fulfilment of: PSA kinetics consistent with biochemical control (Phoenix definition), absence of retreatment/salvage therapy, pad-free continence, maintenance of normal urinary function-defined as the absence of newly developed clinically significant urinary symptoms throughout follow-up, and recovery of erectile function (IIEF-5 ≥ 17 or effective pharmacologic support). Men with ISUP > 2 or receiving neoadjuvant/adjuvant ADT were excluded from model development to preserve a homogeneous risk spectrum for model derivation, leaving 205 evaluable cases. Logistic regression identified predictors. Discrimination, calibration and decision-curve analysis assessed model performance. FOCS-24 was achieved by 30/205 men (14.6%). Younger age and posterior tumor location showed the most consistent pattern with complete recovery. The nomogram incorporated age, PSA density, PI-RADS lesion size, percentage of positive biopsy cores and axial location. The model achieved an AUC of 0.76 with favorable derivation-set calibration metrics. Limitations include low number of composite events and restriction to cryotherapy and HIFU. FOCS-24 provides a rigorous multidomain definition of complete oncologic and functional recovery after FT. External validation across centers and energy platforms is required before clinical implementation.
Pan-cancer Tumour biology
Comment on "Real-world treatment patterns, recurrence, and overall survival of patients with muscle-invasive bladder cancer undergoing radical cystectomy in U.S. oncology practice".
Resumen no disponible en la fuente.
Pan-cancer Tumour biology
Predict. Support. Act? DeepComp and post-operative risk in gastric cancer.
Resumen no disponible en la fuente.
Pan-cancer Precision oncology
Relating transcriptomics to protein abundance reveals self-driven versus interactor-driven proteins.
Statistically modeling the interdependence between transcriptomics and protein abundances remains a persistent challenge in bioinformatics research. Transcriptomic data and proteomic abundances typically display a moderate Pearson's correlation of about 0.5, while in tumor conditions, the correlation may decrease to a much weaker correlation of 0.2. Given that transcriptomic datasets are commonly analyzed due to their ample availability, it is imperative to understand this correlation in greater detail and to be aware of potential deficiencies while conducting research connecting transcriptomic data to protein abundances. Recently, the Clinical Proteome Tumor Analysis Consortium has collected several large-scale proteomic datasets containing transcriptomic and proteomic abundances from hundreds of patients in The Cancer Genome Analysis cohorts. Utilizing these data, a detailed analysis of how protein and transcript abundances correspond for individual genes was presented. As variables of linear regression models, transcript levels of each gene and of their known protein-protein interaction partners, taken from the STRING database, are used. Interestingly, this resulted in two distinct classifications of genes, with a continuum present in between: "self-driven" genes, for which protein abundances are mainly determined by the gene's own transcript level, as well as "interaction-driven" genes, for which transcript levels of other genes possess far greater relevance for its protein abundance than its own transcript level. Notably, the former displays significantly shorter mRNA half-lives than the latter. Namely, this is observed in the ribosome, which is encoded by rather long-lived mRNAs and displays protein abundance levels that are determined by the transcript levels of the other protein components, rather than by its own transcript level. These findings suggest that biomarker interpretation and therapeutic intervention should consider whether a gene is self-driven or interactor-driven, as protein abundance in the latter may be regulated primarily through interaction partners rather than the gene's own transcript level.
Pan-cancer Tumour biology
Evasins: Biology and Therapeutic Potential of Broad-Spectrum Chemokine Inhibitors.
Ticks are haematophagous parasites that rely on uninterrupted bloodmeals for survival and need an effective strategy to suppress the host inflammatory response. Leukocyte recruitment is a chemokine-mediated hallmark of inflammation, which ticks counteract via anti-chemokine proteins in their saliva named evasins. They are a heterogeneous family of glycoproteins that redundantly bind and inhibit chemokines. We provide a detailed overview of their chemokine-binding properties, as well as their therapeutic efficacy in models of inflammation in vivo. Moreover, we discuss the potential challenges posed by using evasins in a therapeutic setting and explore how the development of modified evasins might aid this.
Pan-cancer Tumour biology
Comparative Efficacy of Tirzepatide Versus Semaglutide for Weight Loss in Adults With Overweight or Obesity: A Systematic Review and Meta-Analysis of Head-to-Head Studies.
This systematic review and meta-analysis aimed to compare the efficacy and safety of tirzepatide versus semaglutide for weight reduction in adults with overweight or obesity. We included randomised controlled trials and observational studies comparing tirzepatide and semaglutide with ≥ 24 weeks of follow-up. The primary outcome was percentage weight change from baseline. Secondary outcomes included absolute weight change, weight-loss thresholds, HbA1c and safety outcomes. Ten studies including 41 381 participants were analysed. Tirzepatide was associated with greater percentage weight reduction than semaglutide (MD -4.28 percentage points; 95% CI -5.28 to -3.28; p < 0.00001) and greater absolute weight loss (MD -4.43 kg; 95% CI -5.56 to -3.30; p < 0.00001). Tirzepatide was also associated with a higher likelihood of achieving ≥ 10%, ≥ 15% and ≥ 20% weight loss, with no difference at ≥ 5%. HbA1c reduction was greater with tirzepatide (MD -0.29%; p = 0.0002). Subgroup analyses by study design and type 2 diabetes status yielded consistent findings. There was no significant difference in treatment discontinuation due to adverse events (RR 1.28; p = 0.54), whereas serious adverse events were more frequent with tirzepatide (RR 1.83; p = 0.007). Overall and gastrointestinal adverse events were similar between groups. Tirzepatide was associated with greater weight reduction, greater glycaemic benefit and a higher likelihood of achieving weight-loss thresholds than semaglutide, but with a higher risk of serious adverse events.
Pan-cancer Tumour biology
ATP Release and Purinergic Signaling Potentiate β-Adrenergic Effects on Brown Adipocytes.
<h4>Aim</h4>Brown adipose tissue (BAT) specializes in energy consumption and thermogenesis in response to cold stress. Brown adipocytes (BA) express purinergic receptors, yet little is known about the source of extracellular ATP and the role of purinergic signaling, and the potential interaction with β-adrenergic signaling. We aimed at elucidating whether: BA could release extracellular ATP; ATP/purinergic signaling impacts β-adrenergic responses; and the expression of purinergic receptors in BAT of mice and cultured BA.<h4>Methods</h4>Real-time ATP release was monitored in a mature BA cell model exposed to adrenergic agonists, insulin, glucose, and pharmacological inhibitors. Cellular assays included oxygen consumption, glycolysis, and lipolysis assays, and Ca<sup>2+</sup> imaging. The protein level of ATP-release protein, pannexin-1, was determined. RNA-sequencing was performed on mouse BAT and cultured BA.<h4>Results</h4>We show that β-adrenergic stimulation of BA causes ATP release through pannexin-1 in a cAMP-protein kinase A-dependent manner. ATP release is increased with glucose but decreased with insulin, the latter due to phosphodiesterase 3 activation. β-Adrenergic stimulation increases oxygen consumption, glycolysis, and lipolysis, while apyrase and pannexin-1 inhibitors decrease these effects. Both extracellular ATP and β-adrenergic stimulation induce Ca<sup>2+</sup> signals, albeit with different kinetics. BAT from mice exposed to cold and BA treated with norepinephrine show regulation of several purinergic receptors.<h4>Conclusions</h4>The present study demonstrates that ATP is released from BA via pannexin-1 channels in response to β-adrenergic stimulation, that extracellular ATP, acting through P2 receptor signaling, potentiates β-adrenergic effects on BA function, and that expression of several purinergic receptors changes in response to β-adrenergic stimulation.
Pancreas Precision oncology
Precision Oncology in Hepatopancreatobiliary Cancer Surgery.
Advances in technology have allowed for the characterization of tumors at the genomic, transcriptomic, and proteomic levels. There are well-established targets for biliary tract cancers, with exciting new targets emerging in pancreatic ductal adenocarcinoma and potential targets in hepatocellular carcinoma. Taken together, these data suggest an important role for molecular profiling for personalizing cancer therapy in advanced disease and need for design of novel neoadjuvant studies to leverage these novel therapeutics perioperatively in the surgical patient.
Pancreas Precision oncology
Precision Oncology in Gastrointestinal and Colorectal Cancer Surgery.
Precision medicine is used to treat gastrointestinal malignancies including esophageal, gastric, small bowel, colorectal, and pancreatic cancers. Cutting-edge assays to detect and treat these cancers are active areas of research and will soon become standard of care. Colorectal cancer is a prime example of precision oncology as disease site is no longer the final determinate of treatment. Here, the authors describe how leveraging an understanding of tumor biology translates to individualized patient care using evidence-based practices.
Colorectal Tumour biology
Gender-Specific Cancer-Related Survival Analysis of Robotic Versus Laparoscopic Low Anterior Resection for Rectal Adenocarcinoma: A Multicenter Study.
<h4>Aim</h4>This study compared short- and long-term outcomes of robotic and laparoscopic low anterior resection (LAR) for rectal adenocarcinoma using sex-stratified analyses.<h4>Methods</h4>Patients who underwent elective robotic or laparoscopic LAR for rectal cancer were included. Prospectively maintained database of outcomes was analysed retrospectively. To minimise selection bias, groups were matched.<h4>Results</h4>Of 409 identified patients, 380 met the eligibility criteria (mean age: 61 years, BMI: 26.2 kg/m<sup>2</sup>). In a matched cohort of males (n = 172) and females (n = 114), robotic surgery was associated with significantly shorter hospital stay (p < 0.01). Harvested lymph nodes, margins, morbidity and local recurrence were comparable in males (p > 0.05). Robotic surgery in females was associated with fewer harvested lymph nodes (p = 0.01). Five-year disease-specific survival rates did not significantly differ between approaches in either gender.<h4>Conclusions</h4>Robotics and laparoscopic LAR achieved similar oncologic outcomes. The findings should be interpreted as sex-stratified comparisons rather than evidence of differential effect of surgical approach according to sex.
Pan-cancer Tumour biology
Multiple Retransplantations After BK Polyomavirus–Associated Metastatic Collecting Duct Carcinoma: Evidence for Protective BK-specific T-Cell Immunity
Resumen no disponible en la fuente.
Pancreas Tumour biology
Oncogenic KRAS Regulates Secretion of Extracellular Vesicles and Surface Membrane Charge via Regulation of Phosphatidylserine in Pancreatic Cancer
Resumen no disponible en la fuente.
Pan-cancer Precision oncology
Ferroptosis in kidney disease.
Kidney disease is a growing global public health concern that is associated with high rates of severe morbidity and mortality. Accumulating preclinical and clinical evidence suggests that ferroptosis, an iron-dependent form of regulated cell death characterized by lethal lipid peroxidation, drives the pathogenesis and progression of diverse kidney disorders, including acute kidney injury, chronic kidney disease, renal cell carcinoma and autosomal-dominant polycystic kidney disease. Ferroptosis of renal tubular epithelial cells and intrarenal immune cells engages complex crosstalk with other regulated cell death pathways, amplifying responses and aggravating renal tissue damage. Several biomarkers of early kidney injury, such as the iron regulator NGAL and the phosphatidylserine-binding protein KIM1, are associated with features of ferroptotic cell death, suggesting their potential utility as indirect indicators of early ferroptotic lesions in the kidney. Furthermore, multimodal imaging platforms that integrate optical, magnetic resonance, photoacoustic and radionuclide techniques enable specific in vivo detection of ferroptotic hallmarks, facilitating early diagnosis of kidney lesions. Numerous preclinical studies have identified a diverse range of synthetic small-molecule agents, natural phytochemicals and metabolic modulators that are able to modify ferroptotic signalling. These agents have been shown to mitigate renal parenchymal injury or suppress the proliferation of malignant kidney cells, robustly establishing ferroptosis as a promising therapeutic target for kidney diseases.
Pan-cancer Tumour biology
Monitoring Early Infection Dynamics in Live Spodoptera frugiperda (Sf) 9 Cells via Optical Photothermal Infrared (OPTIR) Spectroscopy
Resumen no disponible en la fuente.
Breast Tumour biology
Harm Laundering in GPT Models: Evidence That Gender Discrimination Is Transformed Rather Than Reduced Across Safety-Trained Generations
Safety evaluations for large language models rely on surface-form classifiers that report declining harm scores across model generations. We provide evidence that this methodology is systematically incomplete: explicit discriminatory content is transformed rather than removed. We call this \emph{harm laundering}. Analysing 450,000 gender-directed completions across 15 models spanning GPT-2 through to GPT-5 (OpenAI GPT lineage; three demographic conditions), we show that sexual violence clusters prevalent in GPT-2 women-directed output disappear by GPT-4, while men-directed completions gain positive representational territory (caregiving, emotional range, ally identity) that women-directed completions do not. The pattern is most visible at GPT-5: Topic~5 (1,997~documents) frames breast cancer as a men's rights debate, while zero equivalent clusters appear in women-directed output. Three independent classifiers score this content as non-toxic. Sentiment scores invert at GPT-4: early models demean women; later models over-correct. Topic diversity in women-directed completions falls 36\% relative to men at the GPT-4 alignment boundary (W/M~$= 0.58$, from $0.91$ at GPT-2). REGARD representational harm disparity correlates with release date ($ρ= +0.55$, $p = .034$) while Detoxify does not ($ρ= -0.23$, $p = .42$): toxicity scores fall as representational harm grows. We formalise harm laundering as a three-criteria test and provide a three-stage detection protocol applicable to any generative model. Within the OpenAI GPT lineage, toxicity score reduction is not a sufficient proxy for harm reduction.
Colorectal AI and pathology
ERCPMP-Gx: Endoscopic Image and Video Dataset for Morphological, Histopathological, and Genomic Characterization of Colorectal Polyposis
Hereditary polyposis syndromes can be precursor lesions to colorectal cancer and are associated with a broad spectrum of extracolonic tumors. Early identification and accurate classification of these syndromes are essential for timely diagnosis, individualized patient management, and targeted surveillance strategies for affected families. However, public endoscopic datasets are largely organized around the individual sporadic polyp, and none links the polyposis phenotype to histopathology and germline findings at the patient level. Here, we present ERCPMP-Gx, an endoscopic, histopathological, and genomic dataset developed to support the application of artificial intelligence (AI) in the recognition, characterization, and classification of colorectal polyposis. Most procedures were performed using the Olympus EVIS X1 system with white-light endoscopy (WLE), narrow-band imaging (NBI), magnifying NBI (M-NBI), and NBI with near focus modes, yielding 160 images and accompanying video clips. Approximately eighty percent of cases represent clinically and/or genetically confirmed hereditary polyposis syndromes (PG), including familial adenomatous polyposis (FAP), Peutz-Jeghers syndrome (PJS), juvenile polyposis syndrome (JPS), and ganglioneuroma syndrome (GNS), while the remaining twenty percent comprise non-hereditary polyps and polyp-mimicking lesions with overlapping morphological features (Non-PG), included to support differential classification. Each released record is linked, where available, to standardized endoscopic annotations, representative histopathology, and clinically reported germline findings, forming an AI-ready, patient-level annotation framework. The dataset is publicly accessible at Mendeley (https://doi.org/10.17632/nzyfc544bx.2). For the latest updates and further information, readers are referred to the DataBioX website: https://databiox.com.
Pancreas Tumour biology
Towards AI-enhanced control: a numerical technique for trajectory smoothing of a parallel robot for pancreatic surgery
The paper presents a numerical approach for the end-effector trajectory smoothing of a parallel robot designed for minimally invasive pancreatic surgery. The approach is tailored for real-time master-slave control architecture and uses a 3D space mouse for command input for velocity control. The trajectory smoothing is achieved by generating S-curves in the end-effector velocity fields, thus controlling the accelerations, which in turn reduces tissue trauma in the minimally invasive procedures. Real-time control is enabled by segmenting the S-curves based on the command inputs from the 3D space mouse. A special case is considered where the acceleration time is constant for all command inputs. Numeric results demonstrate stable transitions (without abrupt changes) in both the end-effector parameter space and in the active joints parameters, thereby validating the proposed approach. Further work aims to test the approach on an experimental model and integrate it into AI-based training modules.
Breast AI and pathology
FCA-Guided Counterfactual Explanations for Multi-Modal Breast Cancer Diagnosis: A Framework Achieving Perfect Validity with Emergent Sparsity
Deep learning models for multi-modal breast cancer diagnosis achieve high predictive accuracy but remain clinically unacceptable without actionable, counterfactual explanations. Attribution-based methods (LIME, SHAP) are categorically inapplicable to this purpose, as they generate no alternative instances and thus cannot be evaluated on counterfactual quality metrics. This investigation provides empirical evidence that FCA-Guided Counterfactual (FCA-CF) framework that uses a Formal Concept Analysis (FCA) concept lattice as a hard structural constraint on counterfactual search, operating over a multi-modal TCGA-BRCA dataset. We benchmark against four genuine counterfactual methods: Wachter-style CF, DiCE, FACE, and NICE, evaluated on 60 benign-predicted TCGA-BRCA instances. The FCA-CF framework achieves Validity = 1.0000 (100% of counterfactuals successfully flip the prediction), Sparsity = 2.37 features changed (best among all valid methods), and Proximity = 0.900 (normalised L2-based, matching NICE as joint best). The classifier achieves Accuracy = 0.980, F1 = 0.976, ROC-AUC = 0.9947. Ablation analysis confirms that the FCA lattice constraint is the primary sparsity driver (removing it increases sparsity by +40%, p < 0.001, Cohen's d = 0.78), while Phase C greedy refinement accounts for the largest individual contribution (+113% sparsity increase when disabled, p < 0.001, d = 5.01). FCA-guided counterfactual generation achieves a clinically important Pareto-dominant outcome; it is simultaneously the sparsest and among the most proximate of all valid methods, with perfect validity. The emergent sparsity property arising from lattice topology rather than numerical penalty terms constitutes a structurally novel contribution to the counterfactual explanation literature.
Pan-cancer Immunotherapy
Improving Sample Efficiency in Peptide-HLA Binding Prediction with Hybrid Quantum-Classical Neural Networks
Peptide-HLA binding prediction is a critical step in neoantigen identification for personalized cancer immunotherapy and holds significant clinical value. However, the training data available for many HLA alleles are extremely limited, which severely constrains the performance of conventional methods on this task. Parameterized quantum circuits are hypothesized to induce inductive biases beneficial for learning from small datasets, yet their application to biological sequence prediction remains underexplored. To address this, we propose a hybrid quantum-classical neural network (HQNN) specifically designed for peptide-HLA binding prediction. HQNN integrates multi-source biological feature encoding with parallel quantum feature extractors and a quantum-enhanced classifier. On two HLA alleles (A*02:01 and B*07:02), HQNN outperforms a parameter-matched classical CNN baseline across all training sizes, with the performance gap widening as training data decreases. Ablation studies confirm the respective contributions of the quantum feature extraction module and the quantum classifier. In noise-aware simulations, performance degrades only mildly, and such degradation is reasonable and acceptable under realistic quantum hardware noise levels. These results suggest that hybrid quantum-classical architectures can provide practical sample-efficiency gains for immunoinformatics tasks in low-data regimes.
Pan-cancer Early detection
FreqDINO++: A Frequency-Guided Multi-Task Routing Vision Foundation Model for Universal Ultrasound Analysis
Ultrasound image analysis plays a crucial role in cancer screening and prenatal diagnosis, yet comprehensive assessment requires jointly addressing tasks such as lesion segmentation and benign-malignant classification. While recent vision foundation models have shown remarkable universal representations, unlocking their potential for ultrasound is bottlenecked by the considerable domain gap from natural images. Existing methods typically fine-tune heavy vision encoders for isolated tasks, incurring substantial computational overhead while overlooking the underlying commonalities across heterogeneous tasks. In this work, we propose FreqDINO++, a frequency-guided multi-task routing vision foundation model for universal ultrasound analysis. We first introduce a Multi-task Routing Adapter (MR-Adapter) to support parameter-efficient integration of task-common and task-specific knowledge, a Frequency-aware Feature Enhancer (F$^2$-Enhancer) is then designed to capture the rich multi-scale frequency characteristics of ultrasound images, and a Task-aligned Collaborative Decoder (TC-Decoder) is devised to promote collaboration between dense and global prediction tasks through global-local token interaction. Extensive experiments on large-scale multi-task and external single-task ultrasound benchmarks demonstrate that FreqDINO++ consistently outperforms strong baselines and recent foundation models across 27 diverse clinical task scenarios, while also showing promising generalization to unseen data. The code is at https://github.com/MingLang-FD/FreqDINO-Plus.
Breast Early detection
Mammography Foundation Models for Opportunistic Prediction of Major Adverse Cardiovascular Events
Cardiovascular disease (CVD) remains the leading cause of death among women, yet cardiovascular risk assessment often relies on clinical variables that may be missing, outdated, or unavailable in routine care. Screening mammography offers an opportunity for opportunistic cardiovascular risk stratification because it is routinely acquired and contains vascular features, including breast arterial calcifications (BAC), that are associated with cardiovascular risk and events. We evaluate whether mammography specific foundation models, originally pretrained for breast cancer-related tasks, can transfer to cardiovascular risk prediction without cardiovascular specific supervision or explicit BAC annotation. We constructed a 5-year major adverse cardiovascular event (MACE) cohort of 22,497 women linked to electronic health record outcomes, including 500 events (2.22% prevalence). The foundation models achieved AUROCs of 0.823 and 0.822 substantially exceeding an age-only model (AUROC 0.765), despite using only the screening mammogram as input, with no clinical variables. Both foundation models evaluated assigned substantially higher predicted risk to patients with radiologist-documented BAC, despite BAC never being used as a training label, and showed activation patterns consistent with vascular findings. Together, these findings suggest that mammography foundation models can recover clinically relevant cardiovascular risk information directly from mammographic pixels and suggest that screening mammography may provide an opportunistic source of cardiovascular risk information to complement conventional clinical assessment without additional imaging. Code is available in https://github.com/PauFeld/MammoCVD
Pan-cancer Tumour biology
"I Know Where to Look," But Does the LLM? Charting the Gaps Between Clinical Expert Needs and Unstructured Data Abstraction Tools
Clinical data abstraction, the process of distilling structured information from patient records, plays a key role in advancing knowledge about diseases such as cancer. Information extraction (IE) with large language models (LLMs) could accelerate this process, but it is unclear whether current frameworks effectively support clinical researchers without AI expertise. To address this, we co-designed an interactive LLM-based abstraction system called Libretto with seven cancer research teams, then evaluated the system's ability to help them answer real-world research questions. We found that while clinicians knew where and how to annotate complex concepts in patient notes, in twelve of fourteen tasks they faced barriers to replicating those intuitions with LLMs. Contextual note reliability judgments, difficulties in steering vibe-coded prompts, and inflexible evaluation strategies necessitated fundamental changes to the IE workflow. Our results highlight open problems for HCI research to bridge the gaps between AI data work tools and clinical users' needs.
Breast Tumour biology
On imaging capabilities of the Imaging and Medical beamline at the Australian Synchrotron
Several propagation-based and analyser-based X-ray phase-contrast imaging experiments with over 100 m between the imaged objects and the detector were carried out at the Imaging and Medical beamline (IMBL) of the Australian Synchrotron. These experiments were aimed at characterization of possible phase-contrast imaging setups at IMBL. Effects of the beam divergence, polychromaticity and X-ray source size on the contrast and spatial resolution of images were evaluated quantitatively in different imaging scenarios. The results of this work will be used for improving the existing and developing new imaging capabilities at the beamline. Applications to X-ray phase-contrast tomography of biomedical samples, particularly for medical breast cancer imaging, are briefly discussed.
Colorectal Tumour biology
A Combined ODE Model of Carbohydrate Fermentation and Colorectal Cancer
We formulate and analyze a system of non-linear ordinary differential equations that describe key metabolic and immunological interactions between butyrate produced by fiber-fermenting gut microbiota, colorectal cancer cells and host cell populations. The model is studied both independently and in conjunction with a pre-existing carbohydrate fermentation model. The parameter space is explored through sensitivity analyses. Simulation experiments are conducted to illustrate the emergence of varying dynamical behaviour driven by butyrate availability. Our model predicts that butyrate production is driven by fiber consumption and further supported by probiotics in the case of microbial dysbiosis. It also suggests that butyrate may help in suppressing tumour growth. We also show that by adding noise with sufficiently high intensity, cancer elimination occurs almost surely in infinite time and that this threshold level of noise intensity decreases with increasing butyrate concentrations.
Pan-cancer Tumour biology
Large Language Model based air quality monitoring and localized alert generation
Poor indoor air quality can cause up to five times more direct health problems to occupants than outdoor air. In particular, it may cause headaches, fatigue, eye/throat irritation, and long-time exposure is linked to respiratory and heart as well as some forms of cancer. Despite the importance of indoor health and well-being, most current monitoring devices and systems (usually for offices and workspaces) are passive. The Environmental Quality Monitor (EnQyMo) platform is a generic Internet of Things (IoT) middleware designed to process several sensor data related to air quality in indoor spaces and correlate this data with health exposure risks of users/workplace employees. Using Bluetooth Low Energy (BLE) beacons and a mobile IoT middleware it is able to identify the (smartphone) users exposed to these polluted air or high CO2 (carbon dioxide) levels, and generate location-specific alarms only to the users at the places with the unhealthy air conditions. At the core of EnQyMo is an agency of Large Language Models (LLMs) capable of interpreting regulatory standards and scientific literature to automatically identify critical health exposure levels.
Pan-cancer Tumour biology
Dense HeLa cell monolayers remain liquid-like despite strong crowding
Collective dynamics in dense cell monolayers are governed by the interplay between crowding and cellular motility. Although increasing density can slow cellular motion and promote glass-like behaviour, the dynamical state of dense HeLa monolayers remains unclear. Here, we combine in vitro time-lapse imaging of HeLa cell monolayers with simulations of a deformable active-cell model to examine how cell density and motility regulate collective relaxation. Within the experimentally accessible density and time ranges, untreated HeLa monolayers remain liquid-like: structural relaxation progressively slows down with increasing density but remains observable throughout the investigated range. Under low-nutrient conditions, cell motility is strongly reduced, and structural relaxation becomes substantially slower. To elucidate the mechanisms underlying these experimental observations, we further performed simulations using a deformable-cell model. The model qualitatively reproduces the density-dependent increase in structural relaxation time and further shows that reducing self-propulsion promotes long-lived caging dynamics at high packing fractions. These results show that dense HeLa monolayers can sustain slow, heterogeneous, yet relaxing collective dynamics under untreated conditions, and indicate that persistent cellular motility is an important factor in maintaining structural relaxation at high density, which may provide insight into the metastatic potential of cancer cells.
Pan-cancer Tumour biology
Doubly valid and doubly sharp sensitivity analysis to unobserved confounding for survival outcomes
Time-to-event outcomes are central in oncology and rare diseases, where treatment effects are often summarized by differences in survival curves or Restricted Mean Survival Time (RMST). In real-world data, estimating these causal effects relies on the absence of unobserved confounding, an assumption that is rarely satisfied. We develop a sensitivity analysis framework for causal treatment effects with survival outcomes under the Marginal Sensitivity Model (MSM). We introduce doubly valid and doubly sharp (DVDS) bounds for differences in survival functions and RMST, extending recent DVDS results to the time-to-event setting while accounting for informative censoring. In practice, our method yields tighter bounds and improved computational efficiency compared to a previous approach from the literature, on simulated and real data. For tractability, we assume independence between censoring and unobserved confounding, a limit that should be addressed in future works.
Breast Tumour biology
A Lightweight CNN Integrated Compact Convolutional Transformer for Multi-Scale Feature Learning and reducing computational complexity for breast cancer mammography image detection and classification
Over the years, Convolutional Neural Networks (CNNs) have demonstrated strong capability in cancer detection and classification using medical images. However, CNN-based models often struggle to capture long-range contextual dependencies. In such scenarios, integrating Compact Convolutional Transformer (CCT) architectures after the CCT layer allows CNN-extracted features to reshape into compact patch tokens using a CCT tokenizer, followed by the addition of positional embeddings to preserve spatial structure. Using 5-fold cross-validation, the model was tested on 3 sets of breast cancer mammography. With only 250,435 parameters, the model achieved 99%-100% accuracy across 3 datasets, indicating robust generalization. Explainable AI (XAI) was integrated into the model to explain the breast cancer classification process to enhance clinical trust. The results indicate that the proposed framework is suitable for computer-aided diagnosis systems, particularly in resource-constrained clinical environments. The novelty of the proposed CNN-integrated CCT overcomes the limitation of CNN's gradient degradation in the last layers by integrating convolutional tokenization with transformer-based learning. Lighter than ViT, which is effective in capturing long-range dependencies, the model has also proven efficient in breast cancer classification by capturing long-range dependencies among breast tissue regions.
Pan-cancer AI and pathology
Semi-Supervised Learning-Based Genetic Biomarkers Dataset for Multiple-Stage Hepatocellular Carcinoma Prediction
Liver cancer is a complex disease responsible for a high number of deaths across the globe each year, making automated solutions for liver cancer classification urgent. The most common form of liver cancer is hepatocellular carcinoma (HCC), accounting for over 90% of liver cancer cases. There is a distinct lack of publicly available HCC datasets utilizing genomic data, which is necessary for training artificial intelligence (AI) models for automated HCC classification. This study proposes constructing a multi-stage HCC dataset using XGBoost and Semi-Supervised learning on three separate datasets of genomic biomarkers, utilizing their existing labels in the Semi-Supervised learning process to label the proposed dataset. The proposed dataset consists of 770 patient samples in total, categorized into five classes that represent normal tissue alongside different stages of HCC. Each sample in the dataset consists of 11,150 different gene expression levels. The XGBoost model demonstrated a final classification accuracy of 96.5% during the Semi-Supervised learning process.
Brain Tumour biology
NeuroTS-Net: Multi-Class Semantic Segmentation of Pediatric Brain Tumors in Multi-Modal MRI
Pediatric brain tumors are a leading cause of cancer-related mortality in children, and their small, rare, and often low-contrast subregions make accurate manual delineation challenging. Reliable automated segmentation is therefore needed to support diagnosis, treatment planning, and response assessment. Accordingly, we introduce NeuroTS-Net, a three-dimensional encoder-decoder convolutional neural network architecture for multi-class semantic segmentation that incorporates a dual-scale raw-detail stream, adaptive low-resolution context selection, and detail-preserving multipath downsampling. These components preserve fine intensity and boundary information while efficiently modeling broader tumor context. NeuroTS-Net was trained on the BraTS 2026 pediatric dataset without external data or pretrained weights and evaluated against nnU-Net and MedNeXt under the same experimental protocol. NeuroTS-Net outperformed the baseline methods, achieving whole-tumor and tumor-core Dice scores of 0.938 and 0.937 on the internal validation set and 0.927 and 0.926 on the official challenge validation set. The code is open-sourced at: https://github.com/maenstru56/NeuroTS.
Breast AI and pathology
GRIN+: Towards Fast Yet Effective Machine Unlearning for Imbalanced Medical Data
As deep learning models become fundamental to modern healthcare, the "Right to be Forgotten" mandated by privacy regulations like GDPR and HIPAA necessitates effective machine unlearning (MU) to remove sensitive patient data from trained models. However, existing MU techniques often struggle with a fundamental "privacy-efficiency-utility" (PEU) trilemma, particularly in medical scenarios where data is frequently characterized by severe class imbalance and long-tailed distributions. In such cases, standard unlearning methods can fail to protect key clinical knowledge or mistakenly delete features essential for diagnosing rare conditions due to the gradient dominance of majority classes. To address these challenges, we propose GRIN+, a novel machine unlearning framework designed for fast and precise data erasure in imbalanced medical scenarios. GRIN+ decouples unlearning-specific knowledge from generalized representations at the parameter level by analyzing the gradient contributions of both "forget" and "retain" sets. It introduces a class-adaptive influence scoring mechanism to rectify gradient dominance and employs a direction-constrained update strategy to prevent the unintended erosion of vital clinical knowledge. Comprehensive benchmarking across multiple medical datasets, including skin cancer (ISIC), brain tumor (MRI), and breast ultrasound (BUSI), demonstrates that GRIN+ achieves an optimal balance of the PEU trilemma. Experimental results show that GRIN+ maintains high diagnostic accuracy and robust privacy while significantly enhancing runtime efficiency compared to existing baselines. We open-source the GRIN+ code and benchmarks to support further research.
Pan-cancer Precision oncology
Democratizing Clinical Tumor Whole Genome Sequencing: 18-hour End-to-end Analysis via Trillion-parameter Large Language Models Locally Deployed on Consumer-grade Hardware
Whole genome sequencing (WGS) is essential for precision oncology, yet its clinical adoption remains limited by prohibitive computational costs and multi-day turnaround times. This work presents a fully localized low-resource framework enabling stable deployment of a trillion-parameter biomedical LLM on a single consumer-grade RTX 4060 laptop with 32GB system memory and 8GB VRAM, as well as on routine clinical workstations in general hospitals, completing the entire tumor-paired WGS workflow from raw FASTQ input to clinical-grade full-variation-spectrum report output. Under standard 30X depth configurations, our implementation finishes a single tumor-paired WGS analysis within 18 hours, achieving 99.62% F1 score for somatic variant detection with over 99.9% concordance to the industrial-standard A100 cluster pipeline, fully meeting clinical oncology accuracy requirements. Quantitative profiling shows adaptive heterogeneous memory scheduling accounts for 71% of total execution time, while model optimization introduces less than 9% of total detection error. This work is the first engineering implementation of trillion-parameter biomedical LLM-driven clinical-grade genomic analysis on consumer-grade hardware, breaking the industry paradigm that trillion-scale genomic LLMs require hundred-thousand-dollar GPU clusters and multi-day turnaround, establishing a low-resource pathway for global primary medical institutions to adopt whole-genome precision oncology at zero additional cost.
Prostate Tumour biology
A Unified Vision-Language Model for PSMA PET/CT Report Generation, Visual Question Answering, and Lesion Segmentation
Accurate PSMA PET/CT interpretation is central to prostate cancer management, yet existing PET/CT AI models typically address isolated tasks. We propose a unified PSMA PET/CT vision-language model for report generation, visual question answering, and lesion segmentation. The framework adopts an LLaVA-style architecture, comprising a PET/CT vision encoder, an MLP-Mixer projection module, a LoRA-tuned large language model, and a 3D segmentation branch. Training followed a four-stage strategy: vision encoder pretraining, projection-layer alignment, VLM fine-tuning, and final multitask tuning. Language tasks used 5,747 PSMA PET/CT datasets with paired reports, while segmentation used the PSMA subset of AutoPET. The model outperformed PET2REP and a CT-based baseline across standard report-generation metrics, improved performance across VQA question types, and achieved higher Dice and lesion-level overlap F1 than SegAnyPET and nnUNet. These results support the feasibility of a unified framework for structured, interactive, interpretable PSMA PET/CT analysis with voxel-level grounding within a single multitask model architecture.
Pan-cancer Tumour biology
HypoEvolve: Genetic Algorithms Enable Multi-Agent LLMs to Discover Scientific Hypotheses
Scientific agents contribute to hypothesis discovery by synthesizing evidence, assessing proposals, and developing new explanations. Recent systems combine scientific agents with evolutionary search through critique, comparison, and revision. However, how different forms of agent collaboration affect hypothesis quality remains an open question. Answering this question requires separating the effects of agents' scientific capabilities from those of their collaboration. A framework must therefore preserve agents' scientific roles and support rules for combining, revising, and retaining hypotheses. Building on this view, we introduce HypoEvolve, which makes collaboration explicit through successive updates to a hypothesis population. Specifically, we propose a generational genetic algorithm to coordinate specialized large language model (LLM) agents that integrate mechanistic arguments, reconsider assumptions, and assess evidence and testability. Each generation specifies how scientific judgments and new proposals reshape the population, making collaboration effects on hypothesis quality directly testable. Moreover, we design our evaluation around scientifically meaningful hypotheses that explain how a proposed intervention could work. Drug repurposing links these explanations to target-level biological claims assessed against external evidence. Specifically, we adapt DepMap and Open Targets into complementary external measures grounded in experimental, genetic, and clinical evidence. Across 34 cancer types, HypoEvolve achieves the highest scores against six baselines on both measures. DepMap selectivity reaches 0.171, versus 0.115 for the strongest baseline. Gains over single-pass generation also generalize to held-out cancer types. HypoEvolve advances a vision of autonomous science in which AI research teams achieve a capacity for discovery beyond that of individual models.
Colorectal AI and pathology
Potential of Artificial Intelligence Algorithms for Identification of Relevant Diagnostic and Prognostic Biomarkers of Early-Stage Liver Cancer
This study explores the use of deep learning and explainable artificial intelligence to diagnose hepatocellular carcinoma (HCC) and define effective biomarkers across five different stages of disease development using a transcriptomic biomarker HCC dataset constructed via semi-supervised learning from three source datasets. Several deep learning experiments were conducted with different feature extraction techniques and gene sets to identify the most effective features for training high-accuracy models with minimal loss. The best-performing model, using 15 selected genes with the SelectKBest algorithm, achieved 90.74% accuracy, while the model with the lowest recorded loss of 0.3187 was obtained using 20 selected genes. To address the issue of class imbalance in the dataset, a weighted training approach was conducted, and for model transparency and interpretability a SHAP-based XAI analysis provided insights into the model's decision-making, consistently finding DNAJB14 as the most influential gene. Functional validation in this study has provided compelling evidence that DNAJB14 plays an important role in the adverse properties of HCC and that its inhibition effectively reverses tumour cell migration, invasion, colony and sphere formation. The main limitation of this study is the dataset's class imbalance, and while weighted training helped mitigate this, further research and additional data are needed to guarantee model generalizability. Future studies should also explore the influence of genetic variations, environmental factors, and clinical differences on model performance across diverse populations.
Pan-cancer Tumour biology
bayprior: Structured Bayesian Prior Elicitation, Conflict Diagnostics, and Regulatory Reporting
Bayesian clinical trials require a well-specified prior distribution, and regulators increasingly expect that specification to be documented, diagnosed for prior-data conflict, and shown to be robust to reasonable alternatives -- yet few R tools connect the full workflow of prior construction, validation, and regulatory justification into a single package. We introduce bayprior, an R package and Shiny application that implements structured expert elicitation via quantile matching, moment matching, and the SHELF roulette method across six distribution families; linear and logarithmic expert opinion pooling with Bhattacharyya agreement diagnostics; prior-data conflict assessment using the Box p-value, surprise index, information divergence, Bhattacharyya overlap, and multivariate Mahalanobis distance; hyperparameter sensitivity analysis with tornado and influence heatmap visualisations; and robust, sceptical, and power prior alternatives. A regulatory reporting module generates self-contained HTML, PDF, or Word documents addressing expectations in the FDA's 2026 draft Bayesian methods guidance and in EMA guidance on incorporating external and historical information. The package is demonstrated on a synthetic oncology Phase II trial. bayprior is available on CRAN and includes a fully modular Shiny application for interactive use.
Pan-cancer Tumour biology
Occupancy Network-Guided Autonomous Robotic Partial Nephrectomy
Autonomous soft-tissue cancer surgery has been limited to interventions on organ surfaces, because current systems cannot perceive and adapt to anatomy once it deforms or is cut. We introduce the first vision-guided autonomous system capable of performing complete tumor resections for partial nephrectomy. Our system integrates conditional occupancy networks, trained entirely in a physics-based simulation, that infer full 3-D anatomy (tumor, margin tissue, and kidney) from single-view partial point clouds. These occupancy networks maintain intraoperative tracking even as tissue is cut and deformed, enabling adaptive planning and execution. The surgical platform combines a depth camera for capturing surface point clouds, dual robotic arms for electrosurgical cutting and vacuum-based tissue manipulation, and an autonomous control strategy for tumor resection. In patient-derived hydrogel phantoms under an open partial nephrectomy setting, the robot performed eight consecutive autonomous tumor resections comprising 77 electrosurgical cuts, with all cuts achieving negative surgical margins and 1.61 $\pm$ 0.48 mm mean absolute margin error. This work demonstrates, for the first time, a foundation for supervised autonomous closed-loop, imaging-driven, margin-negative tumor removal in phantoms.
02 / Method
A transparent retrieval workflow
Every result retains its publication, evidence category and source URL.
- 01Collect
- 02Normalize
- 03Retrieve and rank