BMC Medicine最新文献

Background: Cancer stem cells (CSCs) play a crucial role in breast cancer (BRCA) progression and lymph node metastasis. This study aimed to elucidate how CSCs reshape the immune microenvironment during metastatic dissemination, with a particular focus on macrophage and T-cell regulation.

Methods: A mouse orthotopic BRCA model was established to obtain primary tumor (BRCA_PT) and lymph node metastatic (BRCA_LNMT) tissues. Single-cell RNA sequencing and spatial transcriptomics were used to characterize cellular heterogeneity, marker genes, and intercellular communication. TCGA-BRCA data were analyzed for differential expression, functional enrichment, and immune cell infiltration. In vitro, 4T1-S CSCs were used to assess self-renewal, migration/invasion, ISG15-mediated signaling, and interactions with macrophages and T cells. ELISA, western blotting, sphere formation, colony formation, CCK-8, Transwell, luciferase reporter assays, and ChIP were performed. In vivo, subcutaneous and orthotopic mouse models were used to evaluate the effect of ISG15 on tumor growth and lymph node metastasis.

Results: Bioinformatic analyses revealed an elevated proportion of CSCs in BRCA_LNMT, where CSCs likely induced M2 macrophage polarization through TAM-mediated communication. ISG15 was highly expressed in metastatic tumors and associated with M2 polarization and reduced T-cell activation. In vitro, ISG15 enhanced CSC self-renewal and invasiveness, promoted IL-10-mediated M2 polarization, and upregulated PD-L1 via JAK-STAT signaling to suppress T-cell activity. In vivo, ISG15 silencing significantly inhibited tumor growth and lymph node metastasis.

Conclusion: ISG15 in BRCA CSCs promotes lymph node metastasis by driving M2 macrophage polarization and suppressing T-cell activation, highlighting a critical role for ISG15-mediated immunomodulation and a potential therapeutic target.

Background: Cardiometabolic diseases (CMD) are a leading cause of morbidity and mortality. While both family history (FH) and polygenic risk scores (PRS) are predictive of CMD risk, few studies have systematically evaluated their independent and joint effects. This study aimed to quantify the individual contributions of FH and PRS, as well as their combined impact on CMD risk.

Methods: We conducted an analysis of more than 120,000 adults from the All of Us Research Program with available genotypic, phenotypic, and FH data. CMDs including type 2 diabetes (T2D), obesity, hypertension (HTN), and coronary artery disease (CAD) were ascertained from electronic health records. FH was derived from self-reported survey responses, and family history scores (FHS) were constructed by weighting the number and degree of affected relatives. PRSs were computed using validated multi-ancestry PRS weights from the PGS catalog. Cox proportional hazard regression was used to assess associations of FH, FHS, and PRS independently and jointly with CMD in the overall cohort and stratified by genetic ancestry. We also tested for FHS × PRS interactions and conducted mediation analysis.

Results: Positive FH was significantly associated with increased risk of all CMDs, with the strongest effect observed for obesity (HR 2.07, 95% CI: 2.02-2.12). FHS showed the strongest association with T2D (HR 1.34, 95% CI: 1.32-1.37). Higher PRS values were also associated with elevated disease risk, most strongly for T2D (HR 2.13, 95% CI: 2.07 -2.18). FHS and PRS associations were attenuated in the African ancestry population. A statistically significant interaction between FHS and PRS was observed for obesity (p < 0.001). Individuals with a positive FH and high PRS have greater than 2.5-fold risk of developing CMDs compared to those with negative FH and an intermediate PRS. Mediation analysis indicated that PRS accounted for between 13 and 16% of the total effect of FHS across all traits.

Conclusions: Both FH and PRS are associated with CMD risk and provide complementary but distinct insights into disease risk. PRS adds predictive value beyond FH and partially mediates its effect. Integration of both measures may enhance risk stratification and guide precision prevention strategies.

Background: The growing and pervasive use of social network sites (SNS) has raised concerns about their impact on adolescent mental health during this sensitive developmental phase. Existing longitudinal studies are constrained by methodological limitations and limited exploration of underlying mechanisms. We investigated the longitudinal associations between SNS use and depressive and anxiety symptoms in adolescents and whether sleep mediated these associations.

Methods: We analysed longitudinal data from 2350 adolescents from 31 schools in London, participating in the Study of Cognition, Adolescents, and Mobile Phones (SCAMP). The exposure was self-reported duration of SNS use at baseline (aged 11-12 years). Outcomes were depressive and anxiety symptoms at follow-up, analysed as symptom severity and clinically significant symptoms (aged 13-15 years). The associations between SNS use and depressive and anxiety symptoms were assessed via multi-level ordinal logistic regression (symptom severity) and logistic regression (clinically significant symptoms). The mediation effects of insufficient sleep, sleep onset latency, and sleep disturbance were assessed by mediation analysis.

Results: Compared to 0-30 min per day, more than 3 h per day of SNS use at baseline was associated with higher severity levels of depressive and anxiety symptoms (adjusted odds ratio (OR) = 1.47, 95% CI 1.12, 1.93 and OR = 1.40, 95% CI 1.06, 1.83, respectively) and clinically significant depressive and anxiety symptoms at follow-up (OR = 1.70, 95% CI 1.19, 2.42 and OR = 1.60, 95% CI 1.11, 2.31, respectively). The associations between total and weekend SNS use and depressive symptom severity were stronger in girls than boys. Other associations were similar by gender. Insufficient sleep duration (particularly on weekdays) and sleep onset latency at baseline partly mediated the associations of SNS use and depressive and anxiety symptoms (proportion of mediation ranged between 11.1% and 33.1%). The mediation effects of sleep disturbance were less marked.

Conclusions: In a large longitudinal cohort, we found that SNS use exceeding 3 h per day is associated with increased risks of depressive and anxiety symptoms in adolescents. Findings from mediation analysis suggest that addressing poor sleep hygiene in relation to SNS use might mitigate the negative impact of high SNS use. Our findings may inform the development of early secondary school curricula incorporating digital literacy and sleep hygiene education.

Background: Data on the external validation of current dementia risk prediction models has not yet been systematically synthesised. This systematic review and meta-analysis collated results from three previous reviews to evaluate the predictive discriminative performance of dementia risk models when validated in population-based settings.

Methods: Embase (via Ovid), Medline (via Ovid), Scopus, and Web of Science were searched from inception to June 2022 with an updated search conducted up to November 2024. Included studies (1) had a population-based cohort design; (2) assessed incident late-life (i.e. ≥ 60 years) dementia; and (3) reported predictive performance of at least one dementia risk prediction model in an independent validation sample. Information on study characteristics, dementia outcomes, prediction models (including whether they were fully validated [all original variables available and mapped] or partially validated [one or more variables missing or substituted]), and their discriminative performance were extracted in duplicate. Discrimination, quantified by the area under the receiver operating characteristic curve (AUC) or c-statistic, was pooled across studies using a random-effects model. Models were stratified by validation type: fully versus partially validated.

Results: Thirty-six studies were included. Seventeen studies undertook full validation (14 unique prediction models) and were included in the meta-analysis. Predictor count ranged from one to 57. For all-cause dementia, RADaR showed the highest performance (c-statistic = 0.83, 95%CI: 0.80-0.86; n = 2 validations), followed by eRADAR (c-statistic = 0.81, 95%CI: 0.75-0.85; n = 2 validations). The BDSI model had the most validations (all-cause dementia c-statistic = 0.72, 95%CI: 0.69-0.75; n = 13 validations; and Alzheimer's disease c-statistic = 0.74, 95%CI: 0.61-0.87; n = 2 validations) and performed similarly across high- and middle-income counties. Most validations (76%) were conducted in high-income countries, with 24% in upper-middle income countries. Considerable variation in heterogeneity was observed across models (I² values ranging from 0 to 99%).

Conclusions: Several dementia risk prediction models demonstrate moderate to high external validity. The BDSI model, tested across multiple settings and dementia outcomes, showed promising generalisability. However, the limited number of fully validated models and scarcity of studies in low-income country settings highlight the need for further research on feasibility, resource requirements, and cost-effectiveness before clinical adoption.

Background: Chemotherapeutic agents for ovarian cancer commonly cause chemotherapy-induced peripheral neuropathy (CIPN), significantly impairing quality of life (QoL). Selenium, a potent antioxidant, may mitigate toxicity and improve QoL in cancer patients. This study evaluated intravenous high-dose selenium for preventing neuropathic symptoms in platinum-sensitive recurrent ovarian cancer (PSROC).

Methods: A phase 3, double-blind, parallel group, randomized controlled pilot trial enrolled 68 patients with PSROC, randomized 1:1 to the experimental (selenium) and control (placebo) groups. Patients received sodium selenite pentahydrate (2000 µg /40 mL) or normal saline intravenously two hours before paclitaxel-carboplatin-bevacizumab infusion for six cycles. The primary endpoint was the incidence of grade 1 or more CIPN at 3 months following six cycles of chemotherapy, comparing the experimental group to the control group. Secondary endpoints included comparisons of grade 1 or more, grade 2 or more CIPN before each cycle, 3 weeks and 3 months after six cycles of chemotherapy, adverse events, QoL, and the need for concomitant medications to manage CIPN, and survival between the two groups.

Results: We enrolled sixty-eight patients in the study. The incidence of grade 1 or more CIPN did not differ between the two groups at 3 months post-chemotherapy. However, grade 2 or motor dysfunction incidence was significantly lower in the experimental group before cycle 3 (3.3% vs. 23.3%; P = 0.02) and before cycle 4 (3.3% vs. 20%; P = 0.04), particularly in patients ≥ 60 years. QoL showed no statistically significant difference between the two groups. Duloxetine/gabapentin usage and adverse events were comparable between the two groups, with no selenium-related toxicity, and there were no differences in progression-free and cancer-specific survivals between the two groups.

Conclusions: Intravenous high-dose selenium safely failed to reduce grade 1 or more CIPN, whereas it reduced grade 2 or more motor dysfunction during chemotherapy in patients with PSROC, especially those ≥ 60 years. While the primary endpoint was not met, selenium showed the potential of protective effects against motor neuropathy without safety and survival concerns.

Trial registration: ClinicalTrials.gov Identifier: NCT04201561.

Background: We argue implementation research pays insufficient attention to time. We were prompted by learning gained from the Harnessing Innovation in Global Health for Quality Care (HIGH-Q) programme to explore implementation through time as an analytical lens. Time directly underpins how individuals, teams, and organisations adopt and sustain new practices, yet existing frameworks primarily reference it indirectly. We propose that considering time as a multi-dimensional construct is relevant to the science of implementation in complex systems and to promoting its thoughtful practice.

Arguments: HIGH-Q research involved coordinated ethnographic, quantitative and interventional studies of workforce enhancements in hospitals already benefiting from long-term neonatal technology and quality improvement support. Findings made it clear how time scarcity constrains improvement and use of new technologies in low-resource environments. New clinical technologies such as continuous positive airway pressure require time of users directly and indirectly linked to new cognitive and coordination work. Tasks compete for scarce time resulting in prioritisation, while time is needed for skill development, reflection, and team adaptation. Conceptually we suggest the following: (1) time functions as a finite and negotiable resource that must be deliberately allocated to new practices, without creating temporal space, change efforts risk displacing existing essential work; (2) "hidden time" is required for reflection, collaboration, management and internalisation of new routines-activities rarely acknowledged in project planning; (3) time is an expression of value, reflecting what actors prioritise and the moral or organisational meaning attached to the allocation of effort; (4) healthcare work is governed by temporal structures-shifts, schedules, and social norms-that may hinder flexibility and adaptation; (5) maintaining "time in reserve" supports resilience and psychological recovery in stressful environments, yet interventions may erode this capacity; and (6) implementers' own time investments are frequently omitted when characterising interventions, despite being crucial for sustainability.

Conclusions: Viewing implementation through the prism of time exposes hidden constraints and misalignments between expectations, timelines and real-world conditions. Time in its multiple manifestations should be explicitly examined alongside theories of change and implementation frameworks to help understand why interventions in complex systems succeed or fail, especially where personnel and resources are already scarce.

Background: Spontaneous preterm birth (sPTB) remains a major contributor to neonatal morbidity and mortality, with limited reliable early prediction tools. Existing biomarkers, such as the insulin-like growth factor-binding protein 4 (IBP4) to sex hormone-binding globulin (SHBG) ratio, offer modest predictive performance and are restricted to mid-gestation use (18-20 weeks), limiting their utility for timely intervention. We aimed to develop and validate a novel serological test based on early-gestational sampling to predict the risk of sPTB.

Methods: We conducted a meta-analysis of 18 placental transcriptomic datasets to identify candidate genes associated with sPTB, resulting in 21 protein candidates tested by targeted proteomics. We developed a three-protein panel (glutathione peroxidase 3, GPX3; nidogen-1, NID1; and pappalysin-2, PAPPA2) and validated it in four independent cohorts (456 subjects and 1048 serum specimens) from the USA and Asia. Longitudinal serum samples were collected from 5 weeks and were analyzed using mass spectrometry and ELISA platforms. Predictor performance was compared to the IBP4/SHBG ratio.

Results: The three-protein predictor (GPX3, NID1, and PAPPA2) demonstrated reproducible and superior performance across cohorts: AUC 0.74 (95% CI 0.59-0.88) in Alabama, 0.93 (95% CI 0.88-0.99) in California, 0.80 (95% CI 0.75-0.85) in Asia 1, and 0.83 (95% CI 0.70-0.95) in Asia 2. This outperformed the IBP4/SHBG ratio, which achieved AUCs of 0.68 (95% CI 0.50-0.89), 0.77 (95% CI 0.67-0.88), 0.59 (95% CI 0.52-0.65), and 0.61 (95% CI 0.50-0.75), respectively. Across obstetric trimesters, the three-protein panel maintained high predictive accuracy in the first and second trimesters (AUROC 0.82-0.97), the window when preventive interventions such as progesterone, cerclage, and low-dose aspirin are most effective. Kaplan-Meier analyses confirmed significantly earlier delivery among high-risk pregnancies identified by the three-protein panel.

Conclusions: This maternal serum test provides a reliable approach for early risk assessment of sPTB. The three-protein panel demonstrated reproducible performance across cohorts and across PPROM-positive and PPROM-negative phenotypes, with the strongest discrimination in the first and second trimesters, when preventive therapies such as progesterone or cerclage are most effective. These findings support its potential as an early, clinically actionable screening tool for improving pregnancy outcomes.

Background: Household cannabis use is a risk factor for adolescents' mental health problems. However, little is known about the association of the cessation and psychological impairments in affected adolescents. This study examined the associations of household cannabis cessation and adolescents' mental health outcomes and potential pathways.

Methods: This cohort study used data from the Adolescent Brain Cognitive Development study and included adolescents aged 10-13 years with household cannabis use within 12 months at wave 2. Household cannabis cessation was defined as the absence of cannabis use by household members (excluding the adolescent participant) at wave 3 among households that reported use at wave 2. Internalizing and externalizing problems were assessed using the Child Behavior Checklist, and psychotic-like experiences (PLEs) were evaluated using the Prodromal Questionnaire-Brief Child Version. Family conflict and sleep problems were assessed using the Family Environment subscale and the Sleep Disturbance Scale for Children, respectively. Demographic and psychometric confounders were balanced with propensity score matching (PSM). Linear regression was applied to investigate the associations between cessation and mental health outcomes. Mediation analyses of family conflict and adolescent sleep problems were performed. We further considered the influence of genetic predisposition to cannabis use disorder (CUD) and examined whether brain connectivity patterns, measured by resting-state fMRI, modified the relationships.

Results: Of the 1426 adolescents exposed to household cannabis within 12 months, 438 (30.7%) were no longer exposed by wave 3. After PSM, cessation was associated with lower levels of internalizing and externalizing problems, and PLEs (mean ratios, 0.84-0.86, all P < 0.02), adjusting for baseline scores. The associations persisted after additionally adjusting for the adolescents' polygenic risk for CUD among White participants. Family conflict and sleep problems mediated the associations of cessation with internalizing (proportion mediated, 6.8% and 25.8%, respectively) and externalizing symptoms (14.3% and 24.8%, respectively). Adolescents with weaker connections between cingulo-parietal and dorsal attention networks showed stronger associations between cessation and PLEs.

Conclusions: Household cannabis cessation was linked to a lower level of adolescent mental health problems at follow-up. These findings suggest that interventions aimed at reducing or eliminating household cannabis exposure may be beneficial for youth well-being.

Background: Existing aging clocks, designed to quantify biological aging, primarily capture systemic changes and may overlook alterations crucial for cardiometabolic diseases (CMDs).

Methods: In this study, we developed the CardioMetAge model, an aging clock tailored to predict CMD-related outcomes. Trained in the NHANES-III, the model was applied to the continuous NHANES and UK Biobank. Its associations with cardiometabolic mortality, disease incidence, and transitions between disease states were examined, and its performance in predicting 10-year CMD incidence was also evaluated. We further investigated associations of proteomic pathways, lifestyle factors, and socioeconomic status with CardioMetAge, as well as the impact of caloric restriction intervention on its change.

Results: The final CardioMetAge was constructed as a linear combination of chronological age and 12 common clinical biomarkers. Its age deviation (CardioMetAgeDev) showed stronger associations with CMD mortality (HR per SD [95% CI]: 1.87 [1.83, 1.91]), CMD incidence (1.35 [1.33, 1.37]), and disease progression, including transitions from no CMD to first CMD (1.34 [1.32, 1.35]) and from first CMD to cardiometabolic multimorbidity (1.25 [1.21, 1.30]), compared with deviations of PhenoAge and other traditional biological age models. CardioMetAge also consistently outperformed these models in predicting 10-year CMD incidence. Our findings also highlighted the biological determinants of cardiometabolic aging, with proteomic analyses linking CardioMetAgeDev to inflammatory activation and metabolic disorders. Analysis of modifiable factors revealed that lifestyle and socioeconomic status were associated with CMD risks, partly via CardioMetAgeDev (mediation proportions: 34.5% and 10.7%, respectively). Additionally, two-year caloric restriction slowed the progression of CardioMetAge by 1.23 years (95% CI: [0.61, 1.84]) relative to the ad libitum control.

Conclusions: CardioMetAge outperformed existing aging clocks in ease of use and in predicting CMD-related outcomes. It provides valuable insights into the mechanisms of cardiometabolic aging and holds potential for clinical monitoring and evaluating the effectiveness of interventions.

Background: Opioid use for chronic non-cancer pain remains common in the UK, despite limited evidence of long-term effectiveness. Delirium, a serious acute confusional state associated with increased mortality, is a known adverse effect of opioid use. Pharmacological differences between opioids may influence delirium risk, but comparative evidence is scarce. This study evaluated the association of opioid type and dosage with the risk of in-hospital delirium in non-cancer patients.

Methods: We conducted a retrospective cohort study using electronic health records (EHRs) from a tertiary care hospital in northwest England (September 26, 2014-December 31, 2020). Adult (≥ 18 years) without cancer who were administered with opioids during admission were included. Delirium was identified using the 4 'A's Test or through a combination of ICD-10 codes and new-onset confusion scores (= 3) on the National Early Warning Score. Daily opioid doses were converted to daily morphine milligram equivalents (MME/day) to assess the effect of dose across different opioid types. Incidence rates were calculated by opioid type and opioid dosage. Cox regression models, adjusted for confounders, were used to evaluate delirium risk.

Results: Among 50,586 opioid-exposed patients (mean [SD] age, 55 [20] years; 53% female), 867 patients (1.7%) experienced delirium during their first hospital admission (mean [SD] age, 75.1 [16.7] years). Compared to codeine, oxycodone (hazard ratio [HR] 3.52, 95% CI 2.77-4.46), fentanyl (HR 2.45, 95% CI 1.71-3.51), buprenorphine (HR 2.43, 95% CI 1.54-3.82), combination opioids (HR 2.22, 95% CI 1.63-3.02), and morphine (HR 2.15, 95% CI 1.65-2.79) were associated with significantly higher delirium risk. No clear dose-response association was observed: doses of 50-119 MME/day were not associated with a significant increase in risk compared to < 50 MME/day (HR 0.96, 95% CI 0.66-1.39).

Conclusions: Using in-hospital medication administration records to capture opioid exposure, we found that oxycodone, fentanyl, buprenorphine, morphine, and combination opioids were associated with increased delirium risk compared with codeine. Oxycodone was associated with a higher risk of delirium compared with both codeine and morphine. These findings support personalised opioid prescribing in non-cancer pain and can inform shared clinical decision-making to prevent delirium in patients prescribed opioids.