护理领域文献阅读汇编

Accurate risk stratification in stage II colorectal cancer is essential for treatment decision-making, as current guidelines recommend adjuvant chemotherapy only for patients with a high-risk of relapse. We aimed to develop and validate an AI-based approach for automated invasive front assessment to improve prognostic stratification in this population. We developed SÉMIL (Semantically-Enhanced Multiple Instance Learning), integrating vision-language foundation models with attention-based multiple instance learning for automated invasiveness assessment from H&E-stained whole slide images. We trained and validated SÉMIL on 1,608 H&E-stained WSIs from three cohorts (Austin n=697, MCO n=478, DYNAMIC n=433). We compared SÉMIL performance against manual pathologist assessment and non-semantic MIL approaches. For binary classification, SÉMIL outperformed non-semantic MIL across all cohorts (external validation: AUC 0.713-0.821 vs 0.686-0.803). For survival prediction, SÉMIL demonstrated validated prognostic stratification in both the internal (Austin: HR=4.73, p=0.0012) and the two external (MCO: HR=2.84, p=0.0032; DYNAMIC: HR=2.10, p=0.0396) stage II validation cohorts. Critically, among National Comprehensive Cancer Network (NCCN) guideline-defined high-risk stage II patients, SÉMIL successfully stratified outcomes across all three cohorts (HRs 2.96-3.50, all p<0.05), demonstrating consistent reproducible performance. In multivariate analysis of the combined stage II cohort (n=1,220), SÉMIL retained independent prognostic significance (HR=1.98, p=0.005) after adjusting for conventional clinicopathological features including T stage, MMR status, and lymph node examination adequacy. Concordance analysis between SÉMIL and manual assessment showed concordant infiltrative classification identified the highest-risk group (HR=3.96, p<0.0001), with discordant cases showing intermediate risk. SÉMIL demonstrates validated prognostic stratification in stage II colorectal cancer, with potential utility for refining risk assessment within NCCN guideline-defined high-risk categories where treatment decisions are most challenging.

In TROPION-Breast01 (NCT05104866), datopotamab deruxtecan (Dato-DXd) improved progression-free survival by blinded independent central review versus investigator's choice of chemotherapy (ICC) in patients with inoperable/metastatic hormone receptor-positive, human epidermal growth factor receptor 2-negative (HR+/HER2-) breast cancer, who had disease progression on endocrine therapy and for whom endocrine therapy was unsuitable, and who had received 1-2 prior lines of chemotherapy in the inoperable/metastatic setting. We report detailed safety data and patient-reported outcomes (PROs) from the final analysis. Patients received Dato-DXd (6 mg/kg every 3 weeks) or ICC (eribulin/capecitabine/vinorelbine/gemcitabine). Mouthwash was advised (steroid mouthwash strongly recommended for Dato-DXd), and ophthalmologic assessments were carried out per protocol. Safety and time to deterioration (TTD) in global health status/quality of life (GHS/QoL), pain, and physical functioning, using the European Organisation for Research and Treatment of Cancer QoL Questionnaire-Core 30, were secondary endpoints; exploratory PROs were also assessed. With Dato-DXd (n = 360), nausea was the most common treatment-related adverse event (AE; 51.9%, grade ≥3: 1.4%). The AEs of special interest (grouped terms), treatment-related oral mucositis/stomatitis and ocular surface events, occurred in 57.2% (grade ≥3: 7.2%) and 43.9% (grade ≥3: 1.9%) of patients, respectively. Per protocol recommendations, most patients receiving Dato-DXd received anti-emetics (76.7%), prophylactic mouthwash (77.2%), or artificial tears (71.7%). Adjudicated drug-related interstitial lung disease/pneumonitis with Dato-DXd was mostly low-grade (3.9%; grade ≥3: 0.8%; one grade 5 event). Hematologic toxicity was the most notable feature of the ICC safety profile (n = 351). TTD in GHS/QoL, pain, and physical functioning was delayed with Dato-DXd versus ICC. Patient-reported symptomatic AEs were generally consistent with the clinician-reported AE profile in both arms. Safety and PRO data from TROPION-Breast01 complement the improvements in efficacy demonstrated with Dato-DXd versus ICC, supporting Dato-DXd as a treatment option for patients with previously treated, inoperable/metastatic HR+/HER2- breast cancer.

Cancer immunotherapy, exemplified by immune checkpoint blockade (ICB), remains strongly influenced by the pre-existing immune organization of the tumor microenvironment. Tertiary lymphoid structures (TLSs) are ectopic lymphoid aggregates whose density, maturation state, and spatial localization correlate with clinical prognosis and ICB response across several malignancies. Radiotherapy (RT) can reshape this immune context in opposing directions. By inducing immunogenic cell death, antigen release, cGAS-STING/type I interferon signaling, vascular remodeling, and lymphocyte recruitment, RT can create conditions that support TLS-associated antitumor immunity in selected settings. Conversely, high-dose or large-volume irradiation, poorly timed nodal exposure, and collateral injury to lymphocytes, tumor-draining lymph nodes, stromal scaffolds, and high endothelial venules can disrupt established TLSs or prevent their maturation. This review summarizes current evidence on the bidirectional relationship between RT and TLS biology, differentiating validated mechanisms from indirect evidence and hypothesis-generating translational concepts. We expound on how dose, fractionation, timing, irradiated volume, nodal management and radiation modality may influence antigen presentation, lymphocyte availability, and local immune architecture. Potential strategies such as lymph-node-aware planning, proton or heavy-ion therapy, FLASH RT, vascular normalization, STING or LTβR agonism, and ICB combinations are evaluated as investigational approaches rather than established TLS-directed clinical interventions. Future studies should evaluate whether optimized RT regimens can preserve or promote the functional maturation of TLSs, integrating paired tissue biopsies, spatial transcriptomics, advanced imaging, and circulating biomarkers to definitively correlate TLS remodeling with clinical outcomes.

Therapy resistance is attributed to over 80% of cancer deaths per year, emphasizing the urgent need to overcome this challenge for improved patient outcomes. Despite its widespread use in colorectal cancer (CRC) treatment, resistance to 5-fluorouracil (5FU) remains poorly understood. As an antimetabolite, 5FU imposes substantial metabolic stress, forcing cells that survive treatment to rapidly adapt. We explored acute 5FU-driven changes in mitochondria, the organelle critical for coordinating metabolic stress responses. Here we demonstrate in a range of CRC models that 5FU treatment promotes mitochondrial biogenesis and increases mitochondrial function in surviving cells. Furthermore, we show that targeting mitochondrial metabolism, particularly by inhibiting Complex I, sensitizes CRC cells to 5FU, resulting in delayed tumour growth and prolonged survival in preclinical models. Additionally, analysis of patient data suggests that oxidative metabolism signatures may predict responses to 5FU-based chemotherapy. These findings shed light on mechanisms underlying 5FU resistance and propose a rational strategy for combination therapy in CRC, emphasizing the potential clinical benefit of targeting mitochondrial metabolism to overcome resistance and enhance patient outcomes.

Cardiovascular diseases, particularly stroke, are leading causes of dementia. Several common cardiac interventions such as coronary artery bypass grafting and transcatheter aortic valve implantation are also associated with cognitive decline. However, cognitive outcomes continue to be poorly collected in clinical trials of cardiovascular diseases. In this article, we review the limitations of current approaches to cognitive assessment in cardiovascular disease studies. When assessed, there is wide variation in cognitive tasks used, and tasks have limited population-specific validation, are subject to floor and ceiling effects, and may suffer from sociocultural and linguistic biases. Many tasks are not available in multilingual formats and often rely on face-to-face testing with trained coordinators. All conventional cognitive outcomes in cardiovascular trials are associated with substantial incompletion, especially among older and more impaired individuals, the very people most important to capture. This nonrandom missingness generates survivor and attrition biases, an unacceptable situation for any trial outcome. Last, the meaning of measured cognitive outcomes is often unclear for patients, caregivers, clinicians, and regulators. To help address these limitations, we propose a new framework, major adverse cognitive events (MACE-Cog), to better capture cognitive outcomes in stroke and other cardiovascular populations. As an inclusive construct reflecting the multidimensional nature of cognitive decline, major adverse cognitive events do not rely solely on performance on cognitive testing but also consider inability to complete cognitive testing due to cognitive-behavioral factors, reported symptoms of cognitive decline, impairment in activities of daily living as a result of cognitive impairment, new clinical diagnoses of dementia, and care home admission. We hope that the proposed composite outcome and multiple use cases presented spark progress in cardiovascular research toward more inclusive approaches to the study of cognitive outcomes that move beyond the confines of cognitive tests alone.

Portable high-efficiency particulate air (HEPA) filters effectively remove airborne microbes. To investigate whether HEPA filters reduce respiratory infection episodes in care home residents. This 2-arm cluster randomized clinical trial included care homes for older adults in England, with or without nursing and dementia care provision, and capacity for 20 or more residents in individual bedrooms. Data were collected between September 2021 and May 2024, with each care home participating for 1 winter. Up to 5 HEPA filters for communal areas (clean air delivery rate set at 160 m3/h) and up to 16 filters for bedrooms (clean air delivery rate set at 60 m3/h). Both groups continued usual infection prevention and control measures. The primary outcome was respiratory infection rate per winter per bedroom resident (exposed to bedroom and communal room filters). Secondary outcomes included staff absenteeism. All outcomes were also explored for residents exposed only to communal room filters. During the study period, 91 care homes were randomized, 47 to receive communal room filters, with 569 bedroom residents (median [IQR] age, 87 [81-92] years; 398 [70.0%] female), and 44 without communal room filters, with 589 residents without bedroom filters (median [IQR] age, 88 [82-92] years; 23 [71.8%] female). There was no evidence of a difference in the number of respiratory infections per winter per intervention vs control bedroom residents (0.99 vs 1.04; adjusted incident rate ratio, 0.92; 95% CI, 0.64-1.33; P = .67). There was also no evidence of a difference in the rates of staff absenteeism (adjusted incident rate ratio, 0.80 95% CI, 0.55-1.16; P = .24). Results were similar for the residents exposed only to communal room HEPA filters. In this cluster randomized clinical trial, there was no difference in resident respiratory infection rates in care homes with HEPA filters in communal areas and bedrooms. Care homes should continue existing recommended prevention measures to control respiratory and other infections. isrctn.org Identifier: ISRCTN63437172.

Exercise training improves exercise tolerance (peak oxygen consumption, peakV̇O2) in patients with heart failure with preserved ejection fraction (HFpEF), but it remains unclear whether the effects are comparable between sexes. To evaluate sex-specific exercise training effects on peakV̇O2 in HFpEF. This pooled analysis integrates data from patients with chronic HFpEF randomised to different modes of exercise training or usual care in the OptimEx-Clin and Ex-DHF-trials. Sex differences in changes of peakV̇O2 (absolute, relative, percent-predicted) from baseline to three, six, and 12 months of exercise training or usual care were compared using linear-mixed-models. Among 498 patients (mean age 69.3 ± 7.7 years, 62% women), 277 (57%) were randomised to exercise training. Exercise training significantly increased absolute, relative and percent-predicted peakV̇O2 compared to usual care in women (all P < 0.001), but not in men (all P > 0.05). A significant sex difference in the training response was observed for percent-predicted peakV̇O2 (sex*group*time-interaction P = 0.037). Although mean within-group changes in percent-predicted peakV̇O2 following exercise training were similar between women and men (4.7% [95% CI, 2.2 to 7.2%] vs. 3.0% [-0.3 to 6.4%] at 12 months), usual care was associated with a decline in women (-3.7% [-6.5 to -0.8%]) but not in men (1.4% [-2.0 to 4.7%]). Exercise training improved percent-predicted peakV̇O2 in women but not in men with HFpEF, a sex difference that was largely driven by differing responses following usual care. Since women are usually underrepresented in exercise programmes, more efforts should be made to increase their participation rates. (OptimEx-Clin: NCT02078947; ExDHF: ISRCTN86879094).

Screening for atrial fibrillation (AF) on the basis of AF risk may be more effective. We aimed to develop, externally validate, and prospectively test a machine learning prediction model using electronic health records (EHRs) to guide AF screening. We developed and validated a random forest prediction model for new AF within 6 months, using age, sex, and 10 comorbidities (Future Innovations in Novel Detection of Atrial Fibrillation [FIND-AF] 2.0) in EHRs in the United Kingdom (n=2 081 139), Japan (n=7 795 244), Israel (n=2 166 795), Canada (n=627 919), and China (n=149 145). We conducted a prospective study where participants ≥30 years old without AF and with a CHA2DS2-VASc score ≥2 in men and ≥3 in women, stratified by FIND-AF 2.0 into high and low risk, undertook 4 ECG recordings per day for 3 weeks using a handheld ECG recorder, with a primary outcome of newly diagnosed AF. We estimated stroke risk associated with nonanticoagulated AF in patients with high FIND-AF 2.0 risk in the FinACAF (Finnish Anticoagulation in Atrial Fibrillation) registry of patients with AF (n=229 565). FIND-AF 2.0 was applicable to all EHRs and showed good to excellent prediction performance (United Kingdom: area under the receiver operating characteristic curve [AUROC], 0.819 [95% CI, 0.809-0.829]; Israel: AUROC, 0.835 [95% CI, 0.828-0.842]; Japan: AUROC, 0.751 [95% CI, 0.745-0.757]; Canada: AUROC, 0.747 [95% CI, 0.741-0.753]; China: AUROC, 0.753 [95% CI, 0.725-0.771]), with AUROC>0.7 in men and women in all cohorts, and improved performance compared with CHA2DS2-VASc and C2HEST. Of 1923 participants from 15 sites in the prospective study (mean age, 70.2 [SD 9.4] years), with a mean of 74.8 (SD, 19.4) ECG recordings, AF was diagnosed in 5 of 902 (0.6%) with low FIND-AF 2.0 risk and 46 of 1021 (4.5%) with high FIND-AF 2.0 risk (odds ratio, 8.46 [95% CI, 3.35-21.40], P<0.001). Median AF burden among high FIND-AF 2.0 risk-detected cases was 33.4% (interquartile range, 5.1%-91.6%), and 96.1% initiated oral anticoagulants. In the FinACAF registry, the rate of ischemic stroke for patients with high FIND-AF 2.0 risk, AF, and no anticoagulants was 6.0 events per 100 patient-years. The EHR-based machine learning model, FIND-AF 2.0, identifies a high-risk subpopulation for AF diagnosis among patients at elevated risk of stroke and could enable scalable, EHR-driven, risk-guided AF screening.

Antibiotic-resistant wound infections require dressings that rapidly remove protein-rich exudate, resist fouling and bacterial adhesion, and provide drug-sparing antimicrobial activity. Here we develop asymmetric-interfacial Janus nanofibrous membranes that integrate diode-like exudate transport with infection-microenvironment-responsive nanozyme catalysis for antibacterial wound healing. The membrane couples a multiscale-rough superhydrophobic outer layer with a hydrophilic inner layer, establishing a through-thickness wettability gradient and capillary-pressure asymmetry that preferentially pumps liquids from the hydrophilic side to the superhydrophobic side while suppressing reverse wetting. An asymmetric Janus nanofibrous membrane (Janus-ZnCu-BIF NFM), composed of a hydrophilic nanofiber layer selectively loaded with copper-doped zinc boron imidazolate framework (ZnCu-BIF) nanocubes and an integrated hydrophobic polylactic acid layer, exhibits catalytic antibacterial activity by promoting reactive oxygen species generation and disrupting bacterial membrane integrity. The resulting membrane nearly eliminated viable multidrug-resistant bacteria in vitro. In a murine full-thickness wound model infected with methicillin-resistant Staphylococcus aureus, the dressing significantly reduced bacterial burden and accelerated healing, reaching 99.8% wound closure by day 12. Histological analyses indicated attenuated inflammation, enhanced angiogenesis, and improved dermal regeneration, while transcriptomic profiling revealed enrichment of pathways associated with extracellular-matrix remodeling and vascular development. Overall, this work establishes a scalable, biomimetic dressing that couples rectified exudate management with robust catalytic antibacterial efficacy, providing a practical route toward next-generation wound care materials for drug-resistant infections.

Human oversight of medical AI is increasingly required, but clinician presence alone does not make oversight meaningful. We propose four interlocking conditions—epistemic capacity, cognitive space, decisional authority, and intervention effectiveness—that determine whether human judgment can function as a safety mechanism. Embedding these conditions across procurement, deployment, monitoring, and decommissioning can turn oversight into an operational patient-safety capability across predictive, generative, semi-autonomous, and agentic systems.

Licensed, ready-to-administer injectable medicines can reduce medication errors, minimise waste, and streamline perioperative workflows, but higher acquisition costs have limited uptake in England. This study evaluated the economic impact of switching selected anaesthetic and critical care medicines from conventional ampoules and vials to licensed prefilled syringes within NHS practice. An economic model compared current mixed-use practice with a hypothetical switch to 100% licensed ready-to-administer products for eight medicines: epinephrine 1 mg in 10 ml, ephedrine 30 mg, atropine 3 mg, rocuronium 100 mg in 10 ml, lidocaine (1% and 2%), and midazolam (5 mg in 5 ml and 50 mg in 50 ml). Modelled cost components included medicine acquisition, wastage, nursing preparation time, consumables and preventable adverse drug events. Preparation time reductions were interpreted as capacity release rather than workflow substitution. Deterministic sensitivity analyses explored variation in key assumptions and procurement thresholds. Under the modelling assumptions, epinephrine, ephedrine, atropine and lidocaine 2% were associated with reduced overall system costs of more than £5.3 million annually, largely driven by reduced wastage, preparation workload and modelled adverse drug events. Rocuronium and midazolam were associated with increased costs because of higher acquisition prices despite operational advantages. Sensitivity analyses did not alter the direction of findings. Substantial price reductions would be required for certain medicines to achieve cost neutrality. Licensed ready-to-administer injectable medicines can provide safety and workflow advantages and be associated with economic benefit. Acquisition cost remains a barrier, but broader adoption and market development could improve affordability.

Large language models (LLMs) are increasingly integrated into clinical artificial intelligence (AI) workflows, with emerging applications in medical documentation, information synthesis, guideline retrieval, and patient communication. Recent advances in foundation models, including multimodal systems, have further expanded these capabilities. However, current evidence derives predominantly from examination-style benchmarks, retrospective vignettes, and hospital-based populations, limiting its direct applicability to athlete-centered cardiovascular care. Sports cardiology represents a uniquely challenging environment for AI because of low disease prevalence, physiological-pathological overlap, longitudinal uncertainty, and eligibility-driven decision-making. This narrative review critically examines the conceptual, technical, regulatory, and governance challenges associated with LLM deployment in sports cardiology. We argue that current evaluation paradigms inadequately capture the uncertainty-intensive nature of athlete cardiovascular assessment and propose the text-proxy paradigm, a governance-oriented framework in which LLMs operate as semantic integration layers that process clinician-verified and structured outputs generated by validated upstream systems rather than directly interpreting physiological signals. Within this architecture, LLMs may support documentation, longitudinal synthesis, guideline retrieval, contextual reasoning, and athlete communication under continuous human oversight. We further discuss uncertainty propagation across layered AI systems, workflow-oriented validation, regulatory implications, and a staged framework for responsible clinical implementation. Future progress will depend less on increasingly capable models than on trustworthy clinical ecosystems that prioritize governance, calibration, transparency, auditability, and clinician oversight. Rather than serving as an early adopter of autonomous AI, sports cardiology may provide a benchmark for the responsible integration of foundation models into high-stakes clinical medicine.

The population of Vietnam remains underrepresented in global genomic databases. Here, we present VN1K, a resource of multi-omics and phenotypic information for 1011 unrelated Vietnamese individuals. We present high-depth short-read whole-genome sequencing data for all samples along with various -omics datasets. Using a high-sensitivity variant detection pipeline, which includes a pangenome graph reference and a deep-learning framework, we identify approximately 42 million variants with 7 million short insertions/deletions and 90 thousand structural variants. VN1K also features a whole-genome methylation profile based on long read sequencing. We create a genotype imputation panel with high accuracy on the Vietnamese population, allowing us to identify variants with significantly different allele frequencies in the Vietnamese population compared to other populations. We establish the functional relevance of some of these variants, particularly those in genes associated with genetic disorders, immune diseases, and drug responses, by integrating the allele frequency differences with known genotype-phenotype associations and clinical annotations. Further, we map various loci related to hepatitis B virus infection, triglyceride levels, LDL-C levels, serum glucose levels, HbA1c levels, and levels of two liver enzymes (ALT and AST). The VN1K dataset is accessible via genome.vinbigdata.org, an integrated platform with both linear and graph-based genome browsers.