Citations

Abstract Aims Social isolation (SI) is associated with a higher risk of cardiovascular disease (CVD). One mechanism linking SI and CVD is accelerated biological ageing, which can be assessed using artificial intelligence-enabled electrocardiography (AI-ECG). We investigated whether longitudinal changes in SI status are associated with changes in biological ageing patterns. Methods and results A retrospective longitudinal cohort study including 17 663 individuals without significant CVD who completed ≥2 Berkman and Syme Social Network Index and had paired ECGs within 1 year of each questionnaire. Longitudinal SI status was defined by comparing baseline and follow-up SI status. Ageing trends were assessed using Δage-gap, defined as the change in the gap between AI-ECG-predicted age and chronological age from baseline to follow-up. Linear regression models predicting Δage-gap and survival analyses with Kaplan–Meier and multivariable Cox regression for all-cause mortality were performed. Persistently connected individuals were older [63.1 (52.1, 70.6) years] and had a higher prevalence of hyperlipidaemia and hypertension, yet demonstrated more favourable ageing trajectories. Similarly, those who became connected showed a non-significant attenuation of accelerated ageing. In multivariable linear regression, being persistently isolated [β = +0.412 years, P < 0.001] or becoming isolated [β = +0.438 years, P = 0.012] was associated with accelerated ageing compared to being persistently connected. In survival analyses, becoming isolated (aHR = 1.94, 95% CI = 1.53–2.46, P < 0.001) and being persistently isolated (aHR = 1.66, 95% CI = 1.40–1.96, P < 0.001) were associated with significantly higher all-cause mortality, independent of chronological age and comorbidities. Conclusion Persistent social connection is associated with slower biological ageing and lower mortality. The effects of SI on ageing and survival appear to be potentially reversible, calling for a prospective study to evaluate interventions aimed at improving both the social and biological burden of SI.

Abstract Aims Heart failure (HF), a major global health challenge, affects millions worldwide and poses substantial healthcare and economic burdens. It is estimated that a large proportion of those with early systolic dysfunction remain asymptomatic at a stage when guideline-directed medical therapies have been shown to prevent disease progression. To develop an artificial intelligence (AI) model capable of predicting abnormal left ventricular ejection fraction (EF) directly from static, non-gated, non-contrast chest computed tomography (CT) scans as a form of opportunistic screening, Methods and results Using a multi-institutional dataset of 34 058 paired non- contrast CT images and echocardiogram reports from two academic centres, we trained our model of classification for predicting left-ventricle ejection fraction (LVEF) categories: abnormal EF (EF < 50%) vs. normal on 25 948 studies. We validated the model on 8110 paired chest CT and echocardiogram results from a separate institution. The model achieved an area under the receiver operating characteristic (AUROC) curve of 0.786 on the hold-out test set and 0.762 on external validation to detect an abnormal EF (<50%). Beyond strong predictive performance, the AI model surpassed expert radiologists in both accuracy and efficiency and provided interpretable visualizations highlighting imaging features linked to reduced LVEF. Conclusion In this study, we developed and validated an AI model capable of predicting abnormal LVEF directly from static, non-gated, non-contrast chest CT scans, a novel application for an imaging modality typically used for unrelated indications as a form of opportunistic screening. This technology holds significant promise for early detection of systolic HF, reducing the diagnostic gap, and improving outcomes in asymptomatic HF patients.

Abstract Background Isocitrate dehydrogenase-1 (IDH-1) inhibitor ivosidenib and dual-IDH1/2 inhibitor vorasidenib have demonstrated clinical activity in the treatment of IDH-mutant gliomas. We evaluated the efficacy of vorasidenib in patients with prior exposure to ivosidenib. Methods Retrospective review of patients with IDH-mutant glioma treated with vorasidenib who had prior treatment with ivosidenib. Previous treatments, radiographic response, and adverse events were collected and analyzed descriptively. Results Twenty-nine patients (median age 42, 34% female) were evaluated. Tumor types included IDH-mutant astrocytoma (grade 2 n = 9, 31%; grade 3 n = 3, 10%; grade 4 n = 3, 10%) and oligodendroglioma (grade 2 n = 10, 34%; grade 3 n = 4, 14%). Treatments received prior to ivosidenib included surgical resection (n = 23, 79%), radiation (n = 16, 55%), temozolomide (n = 15, 52%), procarbazine, lomustine, vincristine (PCV, n = 2, 7%), bevacizumab (n = 3, 10%), lomustine (n = 2, 7%), and pembrolizumab (n = 1, 3%). Four patients received additional treatment in between ivosidenib and vorasidenib treatment. Ten of 16 (63%) patients who experienced disease progression on ivosidenib achieved stable disease and remain on vorasidenib, while 6 (37%) had continued progression. Eleven patients with stable disease on ivosidenib transitioned to vorasidenib following FDA approval and remained without progression. Two patients who transitioned to vorasidenib due to ivosidenib-related adverse effects (diarrhea and fatigue) experienced symptom resolution and maintained disease stability. Conclusions Vorasidenib demonstrated clinical efficacy in patients with IDH-mutant gliomas previously treated with ivosidenib. Notably, stable patients who transitioned maintained disease stability, while some of those with progressive disease achieved stabilization with vorasidenib. These findings support vorasidenib as a viable therapeutic option following ivosidenib exposure.

Background: Artificial intelligence (AI), particularly generative and large language models, is being used in nursing education, practice, and scholarly writing. Generative AI applications have been specifically examined for their use in conducting literature reviews with evidence supporting reduced production time of scholarly work. However, there has been limited investigation of their levels of accuracy with identifying references for a literature review. Objective: The purpose of this study was to compare human-generated citations of literature reviews with AI literature-review generated citations. Methods: Using a comparative exploratory design, references from 4 human-written literature reviews, 2 published and 2 unpublished, on 4 different topics, were compared to references derived from 2 AI literature applications, Consensus and Elicit. Three prompting strategies were utilized, including prompts generated using ChatGPT-4. Agreement between the AI and human references was evaluated. Results: The percent of agreement between AI and human generated reference lists ranged from 0% to 63.6%. The Consensus application had a greater overall mean rate of match (21.3%) as compared to Elicit (3.7%). The use of a ChatGPT-4 prompt did not significantly impact results, and there were no differences based on published or unpublished literature reviews. Conclusion: The 2 literature-based applications examined in this study offered a glimpse of their potential use and limitations. The use of an AI literature review application may support but not replace human work.

Artificial intelligence (AI) is rapidly transforming healthcare through applications in clinical decision support, diagnostic imaging, population health management, and workflow optimization. Despite these advances, real-world deployment continues to expose critical challenges related to safety, bias, transparency, and integration into clinical workflows. Algorithmic bias can exacerbate health disparities, limited explainability may undermine clinician trust, and insufficient validation and post-deployment monitoring can compromise patient safety. Although the World Health Organization (WHO) has established six ethical principles for AI in health, including autonomy, well-being and safety, transparency, accountability, equity, and sustainability, translating these high-level principles into practical and enforceable governance mechanisms remains a persistent challenge. This narrative review synthesizes insights from bioethics, health policy, computer science, and clinical medicine to identify gaps in current AI governance approaches and proposes a lifecycle-aligned governance-by-design framework that operationalizes WHO ethical principles across key stages of the healthcare AI lifecycle, including data collection, model development, validation, deployment, and post-deployment monitoring. The framework integrates concrete governance mechanisms such as consent governance, fairness evaluation, external validation, explainability, clinician oversight, and continuous performance monitoring. Overall, this work advances a practical, lifecycle-integrated approach to AI governance and provides a structured foundation for developing safe, equitable, and trustworthy AI systems in healthcare.

Background The optimal imaging guidance for percutaneous coronary intervention (PCI) in complex coronary lesions remains a matter of debate. This meta-analysis compares the efficacy and safety of optical coherence tomography (OCT)-guided PCI versus angiography-guided PCI. Methods Following PRISMA guidelines, we systematically searched PubMed, Embase, Cochrane, and ClinicalTrials.gov from inception to December 2024. Randomized controlled trials (RCTs) comparing OCT-guided PCI to angiography-guided PCI were included. Data analysis was performed using RevMan 5.4, with risk ratios (RR) and 95% confidence intervals (CI) calculated for dichotomous outcomes using a random-effects model. Risk of bias was assessed using the Cochrane RoB 2.0 tool. Results Eleven RCTs involving 6432 participants were included. OCT-guided PCI showed no statistically significant reduction in all-cause mortality compared to angiography-guided PCI (RR 0.72, 95% CI 0.51–1.01, P = 0.06, I 2 = 0%). However, OCT-guided PCI was associated with a significant reduction in cardiac death (RR 0.52, 95% CI 0.30–0.90, P = 0.02, I 2 = 0%), major adverse cardiovascular events (MACE) (RR 0.70, 95% CI 0.55–0.89, P = 0.003, I 2 = 0%), stent thrombosis (RR 0.53, 95% CI 0.33–0.87, P = 0.01, I 2 = 0%), and major edge dissection (RR 0.47, 95% CI 0.34–0.65, P < 0.00001, I 2 = 0%). Upon stratification of studies on the basis of follow-up duration, similar results were noted. Conclusion OCT-guided PCI demonstrated significant benefits in reducing cardiac death, MACE, and procedural complications compared to angiography-guided PCI. These findings support the integration of OCT in complex PCI, although further studies are warranted to assess its impact on long-term mortality.

Background: Resistant hypertension represents a significant clinical challenge, often driven by dysregulated aldosterone production. Aldosterone synthase inhibitors (ASIs) are a novel therapeutic class designed to directly suppress aldosterone biosynthesis. This study aimed to synthesize evidence regarding the comparative efficacy, optimal dosing, and dose–response relationships of three ASIs—baxdrostat, lorundrostat, and osilodrostat—for the treatment of resistant hypertension. Methods: A systematic review and network meta-analysis (NMA) were conducted following PRISMA 2020 guidelines. We searched PubMed, Web of Science, and CENTRAL for randomized controlled trials (RCTs) published up to 31 October 2025. The primary outcome was the mean change in systolic blood pressure (SBP) compared to placebo. We utilized a frequentist random-effects model, characterized dose–response relationships via meta-regression, and ranked treatments using the Surface Under the Cumulative Ranking Curve (SUCRA). Results: The analysis included 8 RCTs comprising 3253 participants. Baxdrostat 2 mg once daily demonstrated the greatest antihypertensive efficacy with a mean SBP reduction of −13.8 mmHg (95% CI: −17.5 to −10.1) versus placebo, followed by Lorundrostat 100 mg (−12.5 mmHg) and Baxdrostat 1 mg (−11.5 mmHg). SUCRA analysis identified Baxdrostat as the highest-ranked treatment (83.0%), followed by Lorundrostat (67.8%) and Osilodrostat (55.5%). Dose–response meta-regression revealed strong linear correlations for Baxdrostat (R2 = 0.91), Osilodrostat (R2 = 0.81), and Lorundrostat (R2 = 0.80). Conclusions: ASIs, particularly Baxdrostat and Lorundrostat, offer robust blood pressure reductions in patients with resistant hypertension. The strong linear dose–response relationships suggest these agents have a favorable therapeutic window for titration. While effective, clinical implementation requires careful monitoring for adverse events.