Daily Respiratory Research Analysis
Analyzed 437 papers and selected 3 impactful papers.
Summary
Today’s strongest respiratory-relevant research spans large-scale clinical artificial intelligence, environmental determinants of asthma, and mechanistic therapeutics for pulmonary hypertension. A multimodal model trained on 3.67 million individuals demonstrated broad cancer prediction capabilities, while complementary studies identified potentially preventable microbial exposures associated with childhood asthma and an AMPK/USP10–ACE2 pathway that may be therapeutically targetable in pulmonary hypertension.
Research Themes
- Multimodal artificial intelligence for clinical prediction
- Environmental microbiome and asthma prevention
- Mitochondrial and endothelial mechanisms in pulmonary hypertension
Selected Articles
1. Advancing cancer detection and treatment using longitudinal routine clinical data.
Oncoformer integrated longitudinal electronic health records and chest radiographs from 3.67 million individuals and 17.7 million clinical visits. It achieved an AUROC of 0.956 for pan-cancer diagnosis, 0.869 for prediction up to one year before diagnosis, mean AUROCs above 0.90 for tumor-stage inference, and validated performance in independent cohorts including the UK Biobank.
Impact: This study advances clinical AI from isolated prediction tasks toward a unified, longitudinal framework spanning diagnosis, staging, treatment response, and recurrence-risk assessment. Its scale, multimodal design, external validation, and use of routine clinical data provide a substantial foundation for future deployment, although prospective implementation remains necessary.
Clinical Implications: The framework could support earlier cancer detection, imaging-assisted staging, individualized treatment planning, and longitudinal surveillance using data already generated in routine care. It should currently be considered a decision-support system rather than an autonomous diagnostic tool, with prospective evaluation of calibration, fairness, workflow integration, and clinical outcomes.
Key Findings
- The model was trained on 3.67 million individuals and 17.7 million clinical visits.
- Pan-cancer diagnosis achieved an AUROC of 0.956, and cancer prediction up to one year before diagnosis achieved an AUROC of 0.869.
- Mean AUROCs for tumor-stage inference exceeded 0.90, and recurrence-free survival stratification was significant across ten cancer types.
Methodological Strengths
- Very large longitudinal multimodal dataset combining electronic health records and chest radiographs.
- Independent external validation including the UK Biobank and validation against postoperative pathological endpoints.
Limitations
- The retrospective nature of routine clinical data may introduce selection, ascertainment, and documentation biases.
- The abstract does not establish prospective clinical utility, causal benefit, or uniform performance across demographic and healthcare-system subgroups.
Future Directions: Prospective, multicenter impact studies should evaluate whether Oncoformer improves diagnostic timeliness, treatment selection, patient outcomes, and health-equity metrics. Future work should also assess calibration drift, interpretability, data governance, and performance across different imaging devices and populations.
Cancer management remains fragmented across its continuum, from late-stage diagnosis and salvage therapies to non-personalized surveillance. Here, we present Oncoformer, a unified multimodal transformer model trained on the China Oncology Multimodal Prediction and Surveillance Study (COMPASS) cohort (3.67 million individuals, 17.7 million clinical visits) and validated on independent external cohorts, including the UK Biobank. Oncoformer integrates longitudinal electronic health records with chest X-ray imaging to address multiple clinical tasks: pan-cancer diagnosis (area under the receiver operating characteristic curve [AUROC] = 0.956), future cancer prediction up to 1 year before diagnosis (AUROC = 0.869), tumor stage inference (mean AUROC > 0.90), patient-specific treatment-response forecasting, and recurrence-free survival stratification across ten cancer types (all p < 0.01).
2. AMPK/USP10 loop activation of ACE2: implications for pulmonary hypertension.
USP10 was reduced in pulmonary endothelium from human idiopathic pulmonary arterial hypertension and rodent pulmonary hypertension. Activation of an AMPK/USP10 positive-feedback loop increased ACE2 Ser-680 phosphorylation and Lys-788 deubiquitination, preserved endothelial ACE2 and pulmonary vascular patency, and attenuated pulmonary hypertension in endothelial USP10-transgenic mice and liraglutide-treated mice.
Impact: This study identifies a previously underappreciated endothelial AMPK/USP10–ACE2 regulatory mechanism and connects it to pulmonary hypertension biology. The use of human disease data, cellular experiments, genetic mouse models, and pharmacological activation provides a strong translational rationale for repurposing or testing glucagon-like peptide-1 receptor agonists, while clinical efficacy remains unproven.
Clinical Implications: The results provide a rationale for investigating GLP-1 receptor agonists, such as liraglutide, as adjunctive or disease-modifying treatments for pulmonary hypertension. Clinical trials are required to determine efficacy, appropriate patient subsets, dose, interactions with approved pulmonary hypertension therapies, and cardiovascular safety.
Key Findings
- USP10 levels were decreased in pulmonary endothelium from human idiopathic pulmonary arterial hypertension and rodent pulmonary hypertension.
- AMPK/USP10 loop activation enhanced ACE2 phosphorylation and deubiquitination, supporting endothelial ACE2 homeostasis and pulmonary vascular patency.
- Endothelial USP10 overexpression and liraglutide treatment both mitigated pulmonary hypertension in mice.
Methodological Strengths
- Mechanistic triangulation across human pulmonary hypertension data, cultured endothelial cells, endothelial cell-specific transgenic mice, and pharmacological intervention.
- The study linked molecular post-translational regulation of ACE2 to hemodynamic and vascular remodeling outcomes.
Limitations
- The evidence is predominantly preclinical, and the abstract does not report a randomized clinical trial in patients with pulmonary hypertension.
- The protective effects of GLP-1 receptor agonism may involve mechanisms beyond AMPK/USP10 activation, and the clinically relevant dose and patient phenotype remain uncertain.
Future Directions: Prospective translational studies should confirm the AMPK/USP10–ACE2 pathway in pulmonary hypertension patients, identify biomarkers of pathway activity, and test GLP-1 receptor agonists in controlled trials alongside standard pulmonary hypertension therapy.
BACKGROUND AND AIMS: The intricate balance between angiotensin-converting enzyme 1 (ACE1) and 2 (ACE2) in the pulmonary vasculature is pivotal for the pathogenesis of pulmonary arterial hypertension (PAH). Catalysing the K48-linked deubiquitination, ubiquitin carboxyl-terminal hydrolase 10 (USP10) is involved in tumour suppression, autophagy, and cell proliferation. This study aims to determine whether a positive feedback loop of USP10 and AMP-activated protein kinase (AMPK) in pulmonary endothelium is protective against PAH. METHODS: In silico data analyses and in vitro culture cell experiments were used to investigate the role of USP10 in human idiopathic PAH (IPAH) and rodent pulmonary hypertension (PH) as well as the underlying mechanism involving a positive feedback loop of AMPK and USP10 in lung endothelium. Endothelial cell (EC)-specific USP10 transgenic (Tg) mice and mice administered liraglutide were used to explore the efficacy of the AMPK/USP10 loop in mitigating PH in rodents.
3. Gram-Positive Bacteria and the Inverse Association between Farm Exposure and Childhood Asthma.
Across two German child populations, nine Gram-positive environmental bacterial genera in farm-associated dust explained approximately two thirds of the inverse association between farm exposure and asthma in mediation analyses. Bioinformatic and mass-spectrometric data linked these bacteria to metabolites acting on the aryl hydrocarbon receptor and peroxisome proliferator-activated receptor-γ, while host receptor polymorphisms modified associations; findings were replicated in independent populations.
Impact: The study moves the farm-exposure hypothesis beyond a broad biodiversity concept toward specific bacterial taxa, metabolites, and host–environment interactions. This creates a mechanistic and potentially actionable framework for microbiome-informed asthma prevention, while the observational design prevents causal conclusions.
Clinical Implications: The findings may guide future development of microbial, metabolite, or environmental-exposure interventions for asthma prevention, particularly in early life. They do not yet justify prescribing specific bacteria or metabolites, because the optimal exposure, safety, durability, and causal effects remain untested.
Key Findings
- A composite score of nine Gram-positive environmental bacterial genera explained approximately two thirds of the inverse association between farm exposure and asthma.
- Metabolic pathways and metabolites associated with the identified bacteria included ligands of the aryl hydrocarbon receptor and PPAR-γ.
- Interactions between the bacterial composite score and human receptor gene polymorphisms were observed, and the principal findings were replicated in independent populations.
Methodological Strengths
- Integrated 16S rRNA sequencing, bioinformatic pathway analysis, mass spectrometry, network analysis, mediation analysis, and gene–environment interaction testing.
- Replication in independent populations strengthened reproducibility and reduced the likelihood that findings were population-specific.
Limitations
- The cross-sectional observational design limits causal inference and permits residual confounding by farm lifestyle, diet, socioeconomic factors, and other exposures.
- The bacterial and metabolite findings require prospective longitudinal validation and controlled intervention studies before preventive recommendations can be made.
Future Directions: Longitudinal birth cohorts and mechanistic intervention trials should determine whether defined bacterial consortia, microbial metabolites, or controlled environmental exposures prevent asthma and identify the developmental window, dose, and host genotypes that determine benefit.
BACKGROUND: The inverse association of farm exposure with asthma and atopic conditions has been attributed to the diversity of environmental bacteria and fungi, but the specific taxa and the underlying mechanisms remain elusive. METHODS: We performed sequencing of the bacterial 16S ribosomal ribonucleic acid (16S rRNA) gene in mattress dust samples from 1018 schoolchildren from two German populations of the cross-sectional GABRIELA (Multidisciplinary Study to Identify the Genetic and Environmental Causes of Asthma in the European Community [GABRIEL] Advanced Study) survey. Bacterial metabolic pathways and enzymes were assessed bioinformatically, and metabolites were measured by using mass spectrometry. Statistical methods included network, interaction, and mediation analyses. RESULTS: Within the highly diverse microbial exposure in a farm environment, the study identified nine gram-positive environmental bacterial genera, whose composite score explained two thirds of the inverse association of farm exposure and asthma in a mediation analysis.