Search Results (15,156)

Cancer cachexia is a multifactorial metabolic syndrome characterized by progressive skeletal muscle loss, affecting 30–60% of patients with advanced renal cell carcinoma (RCC). It significantly impacts treatment tolerance, quality of life, and prognosis, yet its diagnosis and management remain challenging due to fragmented RCC-specific evidence, particularly in the era of immune checkpoint inhibitor (ICI)-based therapy. The pathogenesis involves persistent systemic inflammation, metabolic reprogramming, and tumor–host interactions. The IL-6/STAT3 and TNF-α/NF-κB pathways are central to muscle catabolism, while tumor-derived mediators such as GDF15 and PTHrP, along with mitochondrial dysfunction, further drive cachexia progression. For prognostic assessment, CT-derived skeletal muscle mass evaluation combined with systemic inflammatory and nutritional biomarkers—including neutrophil-to-lymphocyte ratio (NLR), modified Glasgow Prognostic Score (mGPS), prognostic nutritional index (PNI), and cachexia index (CXI)—has improved risk stratification in advanced RCC. Preclinical and emerging clinical data suggest that targeted therapies may partially attenuate cachexia by modulating inflammatory signaling, while multimodal interventions integrating nutritional support and exercise rehabilitation remain the cornerstone of management. Novel strategies, such as inhibition of the GDF15/GFRAL axis, are under active investigation. Future research should prioritize identification of early biomarkers, standardization of cachexia assessment, and prospective evaluation of cachexia-directed interventions in the immunotherapy era. Integrating cachexia assessment into routine practice may ultimately enable personalized treatment and improve long-term outcomes for patients with advanced RCC. Full article

Although robust observational data have long associated physical activity with improved colon cancer survival, the lack of randomized evidence has perpetuated a “causality gap,” confining exercise to supportive lifestyle advice and creating a clinical dilemma for surgeons. This paradigm has been fundamentally reshaped by the landmark phase III CHALLENGE trial. This review synthesizes the clinical, epidemiological, and translational evidence establishing exercise as a validated, disease-modifying intervention. In the CHALLENGE trial, a 3-year structured physical activity program significantly improved both disease-free survival (HR 0.72) and overall survival (HR 0.63) in patients with resected stage III or high-risk stage II colon cancer post-adjuvant chemotherapy, demonstrating a clinically meaningful additive benefit following completion of standard adjuvant chemotherapy. Mechanistically, exercise drives a multi-layered host–tumor reprogramming. Each acute exercise bout triggers transient surges of muscle-derived myokines (e.g., IL-6, SPARC) and immune-cell mobilization, enhancing natural killer and CD8⁺ T cell tumor infiltration. Cumulatively, these repeated acute pulses drive chronic adaptations, including downregulation of the pro-proliferative insulin/IGF-1 axis, attenuation of pro-tumor systemic inflammation, and microenvironmental remodeling at metastatic sites. To bridge the implementation gap, exercise oncology must transition from vague recommendations to structured, biomarker-driven clinical prescriptions defining type, dose, and duration. Mirroring the institutional success of Enhanced Recovery After Surgery (ERAS) protocols, physical activity should be systematically embedded into multidisciplinary, post-adjuvant oncologic care pathways. As the primary coordinators of colorectal cancer care, gastrointestinal surgeons must lead the transformation of exercise from optional supportive care into a validated, disease-modifying standard of care. Full article

Epstein–Barr virus (EBV) is an oncogenic human gammaherpesvirus that is prevalent worldwide. Although over 95% of the world’s population harbors this pathogen, the vast majority of infections are asymptomatic, largely due to immune system involvement. However, in certain situations, such as when individuals are immunocompromised, the immune system can lose control of EBV suppression, leading to oncogenesis. While standard chemotherapeutic regimens can be effective against some EBV-associated cancers, these treatments can be toxic to some patients, and disease relapse and/or drug resistance can develop. Additionally, no human vaccine currently exists for EBV. In recent years, biotherapy has become a promising avenue for the clinical management of EBV-positive cancers. As early detection is often correlated with a more positive prognosis, identification of reliable EBV biomarkers has also come into focus. In this review, we summarize advancements made over the last five years in immunotherapy use for EBV-positive cancers and in EBV biomarker identification to detect both EBV infection and biotherapy response. Full article

Background: Papillary thyroid carcinoma (PTC) is the most common thyroid malignancy, with lymph node metastasis (LNM) being a key predictor of recurrence and poor prognosis. Preoperative detection of LNM remains challenging due to the limitations of imaging modalities, leading to inadequate surgical resection in 20–30% of patients. While immune checkpoint molecules have been implicated in PTC progression, the heterogeneity of CD8⁺ T cell exhaustion subsets and their specific association with LNM remain poorly defined. In this study, we aimed to perform an exploratory characterization of the distinct immune landscape of PTC prone to LNM, with a focus on terminal immune exhaustion, in order to generate hypotheses for improved risk stratification and therapeutic strategies. Methods: Fresh PTC tissues from 40 patients (22 LNM-positive and 18 LNM-negative) were analyzed via flow cytometry (FCM) to quantify immune cell subsets, inflammatory cytokines, and chemokines. Immunohistochemistry (IHC) validated CD45⁺ immune cell infiltration. Transcriptomic and clinical data from 448 PTC patients in The Cancer Genome Atlas (TCGA-PTC) cohort were used for bioinformatic analysis consistent with the observed phenotype, including Gene Set Variation Analysis (GSVA) of terminal exhaustion gene signatures. Results: LNM-positive PTC exhibited a unique inflammatory milieu with significantly elevated IL-6, IL-1ra, CCL5, and IL-9 levels (all p < 0.05) in tumor interstitial fluid. FCM analysis revealed that LNM-positive PTC had increased infiltration of total CD45⁺ immune cells, CD3⁺ T cells, and CD3⁺CD8⁺ T cells (all p < 0.05). Critically, terminally exhausted PD-1^hiTIM-3⁺ CD8⁺ T cells were significantly enriched in LNM-positive PTC (p = 0.022) and positively correlated with extrathyroidal extension (p = 0.044). Additionally, LNM risk was associated with increased CD4⁺ regulatory T (Treg) cell frequency (p = 0.023) and elevated CTLA-4 expression on CD4⁺ T cells (p = 0.047). In TCGA-PTC validation, the terminal exhaustion gene signature was predominantly enriched in LNM-positive (p < 0.0001) and advanced-stage PTC (p < 0.001) and strongly correlated with BRAF mutation (predominantly V600E) (p < 0.0001)—the most common oncogenic driver in aggressive PTC. Conclusions: Our findings suggest a terminal immune exhaustion phenotype (characterized by PD-1^hiTIM-3⁺ CD8⁺ T cells and Treg enrichment) as a potential key feature associated with LNM-prone PTC. This phenotype shows consistency across clinical samples and TCGA datasets, linking BRAF mutation (predominantly V600E) to immune suppression and metastatic potential. These insights provide a novel exploratory immune-based biomarker for LNM risk stratification and support the potential of combining anti-PD-1/TIM-3 therapy with BRAF inhibitors for high-risk PTC, which should be confirmed in future studies. Full article

Drug resistance in cancer remains a major barrier to durable therapeutic benefits and limits the effectiveness of chemotherapy, targeted therapy, and combination treatment in multiple malignancies. Increasing evidence indicates that specific kinesin superfamily proteins contribute to tumor adaptation and therapeutic response in a context-dependent manner through their roles in mitotic regulation, intracellular transport, and stress-response pathways. Aberrant expression of multiple kinesin family members has been documented across diverse cancers and is frequently associated with aggressive clinicopathological features, poor prognosis, and resistance to treatment. However, expression alterations alone do not establish functional dependency, and mechanistic validation is required to distinguish true resistance drivers from adaptive tumor states. In this review, we summarize the classification, biological functions, and abnormal expression patterns of kinesins in cancer; discuss the major mechanisms through which they contribute to drug resistance; and examine strategies for targeting kinesins, including natural-product-derived direct inhibitors, small-molecule inhibitor development, rational combination approaches, and structure-guided and computational optimization strategies. We also evaluate the biomarker potential of kinesin dysregulation and the value of advanced preclinical models for mechanistic and translational investigations. Finally, we highlight the major challenges that hinder clinical translation, including target specificity, compensatory resistance, insufficient biomarker validation, and tumor heterogeneity. Future progress will require integration of functional genomics, multiomics profiling, and mechanism-guided therapeutic strategies to determine when kinesin inhibition represents a clinically actionable approach for resistant malignancies. Full article

Background and Objectives: Treatment-emergent neuroendocrine prostate cancer (t-NEPC) is an aggressive resistance phenotype arising during metastatic castration-resistant prostate cancer (mCRPC). Because metastatic biopsy is not routinely feasible in advanced disease, real-world data on clinically defined t-NEPC/aggressive-variant prostate cancer (AVPC)-like disease remain limited. We aimed to characterize the clinical features, treatment patterns, survival outcomes, and prognostic factors of this clinically defined phenotype. Materials and Methods: We retrospectively reviewed 354 patients with prostate cancer treated at our institution between 2010 and 2020. Seventy-four patients with mCRPC who were clinically identified as having a t-NEPC/AVPC-like phenotype and had a treatment plan for platinum- and/or etoposide-based neuroendocrine-directed systemic therapy were included. Overall survival (OS) and radiographic progression-free survival (rPFS) were estimated using the Kaplan–Meier method, and prognostic factors were evaluated using Cox regression analyses. Results: At clinical identification of the phenotype, the median age was 66.4 years, and visceral metastases were present in 67.6% of patients, most commonly in the liver (55.4%). Median intervals from prostate cancer diagnosis and CRPC onset to clinical identification of the phenotype were 47.5 and 22.7 months, respectively. Neuroendocrine-directed therapy was initiated in 72 patients; platinum–etoposide was the most common first-line regimen (n = 41, 55.4%). Median OS was 4.4 months (95% confidence interval [CI], 2.9–5.8), and median rPFS was 3.5 months (95% CI, 2.7–4.2). In an exploratory, unadjusted analysis, median OS did not differ significantly between doublet chemotherapy and monotherapy (7.0 vs. 3.5 months; p = 0.167). Gleason score ≥9, hemoglobin <12 g/dL, and albumin <3.5 g/dL were independently associated with inferior OS. Conclusions: The clinically defined t-NEPC/AVPC-like phenotype was associated with an aggressive clinical course and poor survival. Earlier recognition, improved biomarker-based diagnostic strategies, and more effective therapeutic approaches are needed. Full article

Ovarian cancer remains difficult to detect before dissemination, and individual circulating biomarkers are unlikely to capture its biological and clinical heterogeneity. Multimodal liquid biopsy offers a potential route forward, but most biomarker combinations are developed empirically, with limited consideration of the molecular relationships among the measured signals. In this Perspective, we propose that intracellular miRNA dysregulation may provide an organizing layer for integrating glycoprotein/glycan remodeling with blood-derived volatile organic compound (VOC) signatures, while corresponding circulating miRNA profiles constitute a distinct measurable biomarker layer that may only partially reflect the underlying intracellular regulatory state. The miRNA–glycosylation axis is supported by direct mechanistic evidence that selected miRNAs regulate glycogenes and pathways controlling fucosylation, sialylation, and other glycan features. By contrast, the proposed connection between miRNA dysregulation and VOC production is indirect and likely mediated through metabolic reprogramming, mitochondrial dysfunction, oxidative stress, ferroptosis, and lipid peroxidation. We evaluate the evidence supporting each biomarker domain, prioritize candidate bridge miRNAs, including members of the miR-200 family and miR-34a, and define the principal boundary conditions of the framework, including causal inference, disease specificity, histological heterogeneity, VOC source ambiguity, temporal variability, and cross-platform standardization. We further outline a translational strategy based on matched biospecimens, mechanism-aware computational integration, experimental perturbation, longitudinal analysis, and independent validation at clinically relevant specificity thresholds. The objective is not to assert a fixed causal chain, but to provide a testable, clinically oriented framework for moving multimodal ovarian cancer diagnostics from statistical combination to biologically informed integration. Full article

Background/Objectives: Oxidative stress, chronic inflammation, disrupted apoptosis, and altered autophagy are biological processes implicated in colorectal tumor development. Morin is a plant-derived flavonoid with antioxidant and anti-inflammatory properties, but its limited solubility and bioavailability may limit its biological activity. This study evaluated the effects of free morin and morin-loaded liposomes (MOR-Lips) on 1,2-dimethylhydrazine (DMH)-associated colonic biochemical, molecular, and histopathological alterations in rats. Methods: Rats were randomly assigned to six groups—vehicle control, MOR, MOR-Lips, DMH, DMH + MOR, and DMH + MOR-Lips—and treated for 10 weeks. Serum and colonic tissues were evaluated for cancer-associated biomarkers (CEA, CA19-9, CA125, HMG-CoA reductase), oxidative stress and antioxidant indices, nitrosative and oxidative DNA-damage markers (MDA, NO, 8-OHdG), inflammatory mediators (TLR4/NF-κB/COX-2, cytokines, MPO), proliferative indices (Ki-67, PCNA), apoptotic and autophagy-related regulators (Bax, caspase-3, p53, cytochrome c, BCL-2, p-AKT, LC3-II, Beclin-1, p62), and histopathological and immunohistochemical changes. Results: DMH exposure was associated with increased CEA, CA19-9, CA125, HMG-CoA reductase, MDA, NO, 8-OHdG, TLR4/NF-κB/COX-2, cytokines, MPO, Ki-67, and PCNA. DMH also reduced NRF2/HO-1 signaling and antioxidant defenses, shifted apoptosis-related markers toward a pro-survival profile, altered autophagy-related markers, and produced marked colonic histopathological abnormalities. Both free MOR and MOR-Lips attenuated several of these DMH-associated alterations, with MOR-Lips generally producing greater effects than free MOR. MOR-LIP treatment was associated with restoration of antioxidant marker profiles, reduced levels of inflammatory and proliferative markers, a shift toward a pro-apoptotic marker profile, partial normalization of autophagy-related markers, and improved colonic histopathological appearance. Conclusions: In DMH-exposed rats, MOR-Lips were associated with more favorable redox, inflammatory, proliferative, apoptosis-related, autophagy-related, and histopathological profiles than free MOR. These findings support further investigation of MOR-Lips as a formulation strategy for improving the biological activity of morin. Because quantitative preneoplastic and neoplastic endpoints were not measured, the results do not establish inhibition of colorectal carcinogenesis or chemopreventive efficacy. Full article

Cancer nanomedicine has generated extensive preclinical evidence of improved drug delivery, pharmacokinetics, and tolerability, yet its clinical impact has often remained modest. This narrative review examines the interconnected biological, pharmacokinetic, manufacturing, regulatory, and clinical factors underlying this translational paradox. A structured literature search was conducted primarily in PubMed and Google Scholar, focusing on studies published between 2022 and 2026 while retaining seminal earlier reports. Major biological barriers include protein corona formation, mononuclear phagocyte system clearance, heterogeneous enhanced permeability and retention, complex tumor microenvironments, and intratumoral heterogeneity. These factors limit circulation, tumor accumulation, tissue penetration, drug release, and interpatient reproducibility. Translation is further constrained by off-target accumulation, uncertain long-term toxicity, non-standardized experimental methods, batch-to-batch variability, scale-up challenges, and fragmented regulatory pathways. Clinical experience shows that successful products are dominated by relatively established platforms and reformulations of known anticancer agents, whereas many actively targeted or structurally complex systems have failed to demonstrate sufficient efficacy or safety. Future progress will require mechanism-driven design, human-relevant preclinical models, harmonized characterization, quality-by-design manufacturing, early regulatory integration, biomarker-guided patient selection, and adaptive clinical trials. Aligning nanoparticle engineering with biological and clinical realities is essential for achieving meaningful patient benefit. Full article

Small, heterogeneous clinical datasets pose a challenge for whole-cohort prediction because clinically meaningful treatment or response patterns may be diluted across biologically diverse patients. We describe and evaluate NetraAI, an interpretable dynamical-systems framework for selective subgroup discovery that uses finite-iteration contraction-inspired dynamics and long-range memory (LRM) to identify stable, outcome-linked Model-Derived Subgroups (MDS). This system can abstain by assigning No Call when a stable subgroup assignment is not supported. A large language model (LLM) Strategist is outlined only as a possible future extension; it is not evaluated here and contributes nothing to the results reported. Foundation and language models asked to perform subgroup discovery directly did not recover the structure the specialized discovery step recovered. We demonstrate this framework across three retrospective clinical trial datasets: Clinical Antipsychotic Trials of Intervention Effectiveness (CATIE) schizophrenia (olanzapine vs. perphenazine comparative treatment-preference benchmark), Canadian Biomarker Integration Network in Depression (CAN-BIND) depression (escitalopram response), and Comprehensive Molecular Characterization of Advanced Pancreatic Ductal Adenocarcinoma for Better Treatment Selection (COMPASS) pancreatic cancer (GnP vs. FOLFIRINOX observational regimen-associated response). The benchmark is not a contest between NetraAI and competing predictors: the same eight downstream methods are evaluated with and without what NetraAI discovered. Given the full feature sets and their own selection procedures, those methods were at or near chance on all three datasets, and blind de novo searches by an independent interaction model and by a pretrained tabular foundation model did not recover an equivalent signature or subpopulation. In internal downstream evaluation, given the discovered variables alone—the same patients, the same classifiers, the full cohort and no abstention of any kind—every one of the eight methods improved on every dataset, 24 of 24 method-dataset comparisons, moving from a raw-feature range of 0.46–0.62 AUC to 0.54–0.78. Restricting further to the subpopulation in which those variables hold improved all eight methods again in CAN-BIND and in COMPASS, 16 of 16 comparisons, reaching 0.66–0.83 and 0.96–1.00, respectively; in CATIE, where the called subgroups are the least outcome-homogeneous of the three, it improved only one of eight. Taking the framework as a whole, 23 of 24 method-dataset combinations improved over the raw-feature baseline. NetraAI abstains on patients without stable subgroup structure, calling 27.7% to 40.4% of each cohort. The contribution demonstrated is therefore subgroup discovery rather than downstream prediction, and its beneficiaries are the conventional methods themselves. In COMPASS, NetraAI identified a three-SNV signature associated with regimen-linked response ranking among called patients; because the cohort was observational and the permutation test was conditional on the selected signature, this finding is exploratory. The two mechanisms are complementary rather than competing: variable discovery establishes which features carry the structure, and abstention establishes in which patients it holds. Neither mechanism replaces conventional modeling. Variable discovery improved every method on every dataset, the pretrained tabular foundation model included; identifying the population in which those variables hold conferred further benefit in two of the three datasets and not in the third. These findings support NetraAI as an exploratory system for generating compact, inspectable subgroup hypotheses that may inform future enrichment strategies after external validation, and indicate that its value lies in what it contributes to other methods rather than in competing with them. Full article

Distal-less homeobox 1 (DLX1) has emerged as a promising urinary biomarker for prostate cancer. This study developed a label-free piezoelectric acoustic biosensor for selective DLX1 detection using a quartz crystal microbalance (QCM). The QCM gold electrode was sequentially functionalized with L-cysteine, graphene oxide (GO), and an amine-terminated DLX1 capture probe covalently immobilized through EDC–NHS-mediated amide bond formation. Stepwise surface functionalization was characterized by X-ray photoelectron spectroscopy (XPS), supported by contact angle measurements, X-ray diffraction, and field-emission scanning electron microscopy. XPS provided multi-element evidence for Au–S thiolate formation, GO deposition, amide coupling, probe immobilization, and Watson–Crick hybridization with the synthetic DLX1 target. Under optimized conditions, the biosensor exhibited a linear response to DLX1 concentrations and achieved a limit of detection of 81.19 nM. Non-complementary sequences, including PCA3 and SARS-CoV-2, produced frequency shifts below 9 Hz, confirming high selectivity. Comparative experiments showed that GO-mediated covalent immobilization was essential for reliable detection, whereas direct DNA attachment to bare Au generated anomalous positive frequency shifts, which were attributed to weak physisorption. The proposed platform offers a sensitive and selective strategy for quantitative DLX1 detection and may support future point-of-care nucleic acid diagnostics. Full article

Background/Objectives: Hexavalent chromium [Cr(VI)] is a toxic metal that enters the environment due to natural and anthropogenic processes. The liver is a major target organ of Cr(VI) exposure through drinking water, resulting in an increased risk of exposed individuals developing chronic liver diseases and cancers. Cr(VI) exposure is also associated with neurological dysfunction. Essential metal dyshomeostasis is a well-characterized biomarker of liver and neurological diseases, but the role Cr(VI) plays in altering essential metal levels in both organs is poorly understood. Methods: Male and female Hartley guinea pigs were exposed to 0 or 5 mg Cr(VI)/L in drinking water for 90 days. Chromium accumulation and essential metal levels in the liver, blood and brain were analyzed using inductively coupled plasma mass spectrometry. Results: Chromium significantly accumulated in a sex-dependent manner in the livers and blood of guinea pigs given 5 mg Cr(VI)/L in their drinking water, with sex- and region-specific chromium deposition observed in the brain. The livers and blood of female guinea pigs were more vulnerable to Cr(VI) exposure, as evident in the significant chromium accumulation and dyshomeostasis of several essential metals, including hepatic magnesium, zinc and potassium. Conversely, male guinea pigs were more vulnerable to Cr(VI)-induced essential metal dyshomeostasis in the brain, with significant changes in molybdenum and cobalt in nearly half of the 11 brain regions assessed. Conclusions: This study begins the process of establishing a novel model to study Cr(VI)-induced hepatotoxicity, neurotoxicity and organ crosstalk. Overall, Cr(VI) induced sex-dependent essential metal dyshomeostasis in the liver and blood, along with sex- and region-specific essential metal dyshomeostasis in the brain. These results suggest essential metal dyshomeostasis may play an important role in Cr(VI)-induced liver and neurodegenerative diseases, and these effects should guide future mechanistic investigations. Full article

Although expression-based omics has greatly contributed to biological research, the biochemical state of cells or tissues cannot be fully explained by transcript or gene expression data alone. This systematic review investigated when matrix-assisted laser desorption/ionisation (MALDI)-based approaches can support the integration of multiple molecular layers and when they remain limited to single-class molecular mapping. PubMed was searched between 1 February and 8 June 2026. English-language peer-reviewed publications relevant to the principles, methods, molecular classes, or biomedical applications of matrix-assisted laser desorption/ionisation-based multiomics were eligible. Publications outside this scope were excluded. Records were independently screened by two reviewers, and disagreements were resolved through discussion. The findings were narratively synthesised based on the analytical platform, molecular layer, integrated spatial multiomics, spatial application, and biomedical use. The methodological quality and risk of bias were not formally assessed, which limits the certainty and strength of the conclusions drawn from this narrative synthesis. Following the reapplication of the relevance criteria, 111 publications from 1633 identified records were included in the final review. The evidence was reclassified and presented separately as true multilayer MALDI studies, MALDI integrated with orthogonal modalities, and background uniomics or spatial molecular profiling studies. These publications included human, animal, cellular, tissue-based, and methodological studies. These publications cover many molecular classes and biomedical applications. As many publications did not involve human participants, a single pooled participant total was not applicable to this study. Matrix-assisted laser desorption/ionisation-based techniques were used to analyse metabolites, lipids, glycans, peptides, and intact proteins in the reviewed literature. The literature also shows that spatial information can be retained using matrix-assisted laser desorption/ionisation imaging mass spectrometry (MALDI-IMS). Its ability to map molecular signals back to specific tissue regions or areas associated with disease supports its potential clinical utility. Specific areas of application, such as cancer, neurological, and infectious disease research, are also reviewed. The use of matrix-assisted laser desorption/ionisation-based techniques for biomarker discovery has been evaluated. Small sample sizes, proof-of-concept designs, inconsistent outcome reporting, and limited external validation have limited the evidence. Technical limitations include ion suppression, limited quantification, sample preparation variability, and differing analyte sensitivity. Emerging developments, such as single-cell imaging and AI analysis, have also been explored. Matrix-assisted laser desorption/ionisation enables multiomics when two or more complementary molecular layers are analysed and integrated within the same biological system. When only a single molecular class is analysed, the approach remains spatial molecular mapping rather than multiomic analysis. However, before matrix-assisted laser desorption/ionisation-based approaches can be applied in routine clinical practice, standardised workflows and better quantitative methods are needed. This review received no external funding and was not registered. Full article

Background: Skeletal muscle depletion predicts poor outcomes in gastrointestinal cancers, but whether volumetric three-dimensional (3D) body composition adds anything over the standard two-dimensional (2D) single-slice approach in colorectal cancer (CRC) patients receiving chemoradiotherapy is unclear. We quantified CT-derived sarcopenia and directly compared 2D and 3D body composition metrics from a fully automated deep learning pipeline. Methods: We retrospectively analyzed 368 patients with CRC. Body composition was quantified automatically from CT using SMAT-BC, a pipeline combining TotalSegmentator-based vertebral localization with an nnU-Net Residual Encoder XL network incorporating a Transformer bottleneck for four-class tissue segmentation. Single-slice L3 (2D) and L1–L5 volumetric (3D) indices were derived. The primary endpoint was overall survival (OS); recurrence-free survival (RFS) was secondary. Cox proportional hazards models with bootstrap optimism correction were used. Measurement reproducibility was assessed in 25 external TCGA-COAD cases. Results: Sarcopenia was strongly associated with both overall and recurrence-free survival (unadjusted OS HR 2.16, 95% CI 1.55–3.00; unadjusted RFS HR 1.83, 95% CI 1.36–2.47; both raw p < 0.001; adjusted OS HR 1.89, 95% CI 1.57–2.28; adjusted RFS HR 1.44, 95% CI 1.22–1.70; FDR-adjusted p < 0.0001 for both). L3 single-slice indices were strongly correlated with volumetric indices (r = 0.91 for muscle index) and added discriminant value over the clinical model for overall survival (optimism-corrected C-index: clinical 0.602 (95% CI 0.578–0.626), clinical + 2D 0.631 (95% CI 0.609–0.654), clinical + 3D 0.599 (95% CI 0.574–0.624); DeLong p = 0.002 for clinical + 2D vs. clinical, FDR-adjusted p = 0.006). The 2D- and 3D-augmented models yielded overlapping bootstrap confidence intervals and were not clinically meaningfully different in this cohort (OS ΔC = +0.032, 95% CI +0.013 to +0.051; FDR-adjusted p = 0.006; RFS ΔC = +0.008, 95% CI −0.011 to +0.027; FDR-adjusted p = 0.612). Conclusions: Automated CT-derived sarcopenia is an independent predictor of survival in CRC patients receiving chemoradiotherapy. In our cohort, single-slice L3 measurement matched or exceeded volumetric discrimination, but the 2D- and 3D-augmented models yielded overlapping bootstrap confidence intervals for both endpoints: for overall survival, the DeLong FDR-adjusted p value for the 2D-versus-3D contrast was 0.006, and 0.612 for recurrence-free survival. Because no non-inferiority margin was pre-specified, the 2D–3D comparison is presented as exploratory, and we make no formal claim of non-inferiority or equivalence for either endpoint. The findings support single-slice L3 measurement as an efficient biomarker for risk stratification but warrant external validation in larger prospectively designed cohorts. Full article

Background: Stage II/IIIA colorectal cancer (CRC) patients have a relatively high 5-year overall survival rate after surgical resection alone, but more than 50% of patients receive adjuvant chemotherapy. This study investigates a technology that applies artificial intelligence (AI) to conventional histopathology images to identify patients at risk of recurrence. Methods: HistoNav, a novel AI deep-learning algorithm based on Vision Transformer, Graph Neural Networks and Convolutional Neural Networks was designed to analyse H&E-stained formalin-fixed paraffin-embedded (FFPE) samples (n = 2095) to stratify Stage II/IIIA CRC patients into low-, intermediate-, and high-risk groups. Results: Data analysis revealed a 5-year disease-specific survival (DSS) of 93.2%, 84.3%, and 69.3% for low-, intermediate-, and high-risk groups, respectively. The hazard ratio for the high versus low-risk group (HR = 4.611, 95% CI: 2.776–7.662; p < 0.000001) was statistically significant, demonstrating HistoNav’s potential to stratify patients based on recurrence risk. Conclusions: HistoNav can effectively identify CRC patients with good prognosis using the digital images of H&E-stained resection samples and will support clinical decisions around the use of adjuvant chemotherapy. Full article