Daily Respiratory Research Analysis
Dynamic monitoring of pneumonia severity using a new machine-learning trend score outperformed existing models and generalized across international cohorts. Serum IgA biology emerged as a contributor to emphysema and airway disease via population cohorts and Mendelian randomization. A first-in-class sulfonium lipid nanoparticle enabled intranasal mRNA delivery and genome editing across key lung epithelial and immune cells in mice, opening a path for inhaled genetic therapies.
Summary
Dynamic monitoring of pneumonia severity using a new machine-learning trend score outperformed existing models and generalized across international cohorts. Serum IgA biology emerged as a contributor to emphysema and airway disease via population cohorts and Mendelian randomization. A first-in-class sulfonium lipid nanoparticle enabled intranasal mRNA delivery and genome editing across key lung epithelial and immune cells in mice, opening a path for inhaled genetic therapies.
Research Themes
- Dynamic risk stratification and prognosis in community-acquired pneumonia
- Humoral immunity (IgA) in emphysema and airway remodeling
- Intranasal mRNA delivery and genome editing in the lung
Selected Articles
1. Dynamically monitoring pneumonia severity scores to predict the prognosis in patients with community-acquired pneumonia: an international, multicenter cohort study.
A new trend-based algorithm (Mortality Vector Optimization) quantified changes in CURB-65/PSI between admission and 72 hours, improving in-hospital mortality prediction versus nine ML comparators and generalizing to two external datasets. The derived trend score strongly captured disease progression (HR ~2.5), supporting dynamic monitoring of treatment response in CAP.
Impact: Introduces a practical, externally validated method to dynamically track CAP severity and improve mortality prediction using familiar scores. This can reshape early treatment decisions and escalation strategies.
Clinical Implications: Embedding MVO-derived trend scores into electronic health records could enable real-time risk updates to guide ICU triage, antibiotic de-escalation/escalation, and response-adaptive care pathways during the first 72 hours.
Key Findings
- MVO outperformed nine established ML models for mortality prediction using PSI/CURB-65 dynamics (derivation AUC for PSI 0.825; external AUC up to 0.870).
- Trend score captured disease progression with strong associations to in-hospital mortality (HR 2.63 for CURB-65 trend; HR 2.50 for PSI trend).
- External validation across GOSSIS-1-eICU and SCRIPT CarpeDiem datasets demonstrated generalizability.
Methodological Strengths
- International, multicenter cohorts with predefined timepoints (admission and 72-hour reassessment).
- External validation on independent critical care datasets; discrimination and calibration assessed.
Limitations
- Observational design with potential residual confounding and practice variability.
- Prospective impact on clinical decision-making and outcomes not yet tested in randomized implementation studies.
Future Directions: Prospective, cluster-randomized evaluations of MVO-guided care pathways; integration with EHRs; extension to sepsis/viral pneumonia and incorporation of biomarkers and imaging.
BACKGROUND: Traditional pneumonia severity scores are widely used at treatment initiation to stratify disease severity and predict prognosis. Although they can be reassessed during treatment to evaluate disease progression, their predictive accuracy remains suboptimal. To address this limitation, we introduce a novel machine learning algorithm that quantifies continuous disease progression, enhancing real-time evaluation of treatment efficacy and outcome prediction in community-acquired pneumonia (CAP). METHODS: We analyzed 2052 CAP patients from China and 1862 from Japan. CURB-65 and pneumonia severity index (PSI) scores were recorded at admission and 72 h post-treatment. To quantify disease progression, we developed a new algorithm, Mortality Vector Optimization (MVO), which measures the similarity between a patient's condition and critical states within severity scoring systems. Model discrimination was assessed using the C-index and AUC, while calibration was evaluated with Brier scores. External validation was conducted using the GOSSIS-1-eICU and SCRIPT CarpeDiem datasets. RESULTS: MVO demonstrated superior predictive performance compared to nine established machine learning models. In the derivation cohort, MVO achieved a C-index of 0.734 and AUC of 0.825 for PSI, while in the GOSSIS-1-eICU dataset, it reached a C-index of 0.850 and AUC of 0.870. The trend score derived from MVO effectively captured disease progression, with hazard ratios of 2.63 (for CURB-65) and 2.50 (for PSI) in predicting in-hospital mortality. CONCLUSION: The trend score provides a dynamic and reliable method for monitoring early treatment effectiveness in CAP. MVO offers enhanced predictive accuracy and holds potential for broader applications in disease prognosis assessment.
2. Serum IgA isotypes are associated with percent emphysema, wall thickness and lung function decline.
In MESA (n=5,497), lower serum IgA was associated with higher CT percent emphysema, supported by Mendelian randomization. Higher galactose-deficient IgA1 correlated with airway wall thickening and accelerated decline in FEV1 and FEV1/FVC, with replication in SPIROMICS, implicating humoral immunity in emphysema and airway remodeling.
Impact: Links systemic IgA biology to structural lung disease using imaging, longitudinal function, and genetic instruments, suggesting a potentially modifiable immune pathway in emphysema.
Clinical Implications: IgA-related biomarkers could aid risk stratification for emphysema and airway disease; interventions that enhance mucosal IgA or reduce pathogenic Gd-IgA1 may be explored to modify disease progression.
Key Findings
- Lower serum IgA associated with greater CT-derived percent emphysema (β=-0.084; p=0.005), confirmed by Mendelian randomization (β=-0.79; p=0.011).
- Higher serum galactose-deficient IgA1 linked to increased airway wall thickness (β=0.0079; p=0.012) and accelerated decline in FEV1 and FEV1/FVC.
- Findings replicated across cohorts (SPIROMICS), implicating humoral immunity in emphysema pathogenesis and airway remodeling.
Methodological Strengths
- Large population-based cohort with CT phenotyping and spirometry, plus replication.
- Use of Mendelian randomization to strengthen causal inference.
Limitations
- Serum IgA may not fully reflect mucosal IgA; residual confounding possible.
- Effect sizes modest; interventional relevance requires testing.
Future Directions: Mechanistic studies of mucosal IgA and Gd-IgA1 in lung tissue; trials testing strategies to augment protective IgA or reduce Gd-IgA1; integration with microbiome and infection phenotypes.
RATIONALE: Immunoglobulin A (IgA) deficiency, a rare, highly heritable trait, is associated with frequent pulmonary infections, emphysema, airway changes and low lung function; however, it is unclear if reduced IgA levels may affect lung structure and function. METHODS: Serum IgA, IgA1 and galactose-deficient IgA1 (Gd-IgA1) levels were measured in the population-based Multi-Ethnic Study on Atherosclerosis (MESA). The MESA Lung Study measured percent emphysema on cardiac CT and airway dimensions on chest CT, and performed spirometry. Regression models were evaluated after adjustment for demographic and CT factors. Mendelian randomisation (MR) analyses were conducted using genetic variants from the Trans-Omics for Precision Medicine (TOPMed) programme. A replication analysis was performed in the SubPopulations and InteRmediate Outcome Measures In COPD Study (SPIROMICS). MEASUREMENTS AND MAIN RESULTS: Among 5497 participants, lower log-normalised serum IgA levels were associated with greater percent emphysema (β=-0.084; 95% CI -0.14 to -0.026; p=0.005), which was confirmed on MR (β=-0.79; 95% CI -1.4 to -0.18; p=0.011). Greater log-normalised serum Gd-IgA1 levels were associated with airway wall thickness (β=0.0079; 95% CI 0.0017 to 0.014; p=0.012; n=2580) and decline in the forced expiratory volume in one second (FEV1) (β=-0.012; 95% CI -0.021 to -0.0036; p=0.0055; n=2778) and FEV1/forced vital capacity (FVC) ratio (β=-0.0028; 95% CI -0.0048 to -0.00084; p=0.0054; n=2778). CONCLUSION: Lower serum IgA levels were associated with greater percent emphysema. Additionally, higher Gd-IgA1 levels were associated with airway wall thickness and lung function decline. These findings support a protective role of IgA in emphysema pathogenesis and possible deleterious role of Gd-IgA1 in airway diseases.
3. Sulfonium lipid nanoparticles for intranasal mRNA delivery to lung epithelial and immune cells.
A first-in-class sulfonium lipid nanoparticle enabled intranasal delivery of mRNA to club and ciliated epithelial cells and macrophages in mice, supporting applications from bioluminescence imaging to CRISPR genome editing and cytokine delivery. No sustained toxicity was observed, suggesting a safe alternative to amine-based pulmonary lipids.
Impact: Demonstrates a new chemical class for pulmonary mRNA delivery with in vivo efficacy across multiple lung cell types and payloads, addressing a key translational bottleneck for inhaled genetic therapies.
Clinical Implications: If translated, sLNPs could support inhaled mRNA therapeutics for cystic fibrosis, influenza/RSV prophylaxis, gene editing of monogenic lung diseases, and localized immunomodulation with reduced systemic exposure.
Key Findings
- Sulfonium lipid nanoparticles achieved efficient intranasal mRNA delivery to club cells, ciliated epithelial cells, and macrophages in vivo.
- Enabled functional payloads including CRISPR-Cas9 mRNA/sgRNA for genome editing and cytokine mRNA for immunomodulation.
- Demonstrated safety in adult mice without sustained local or systemic toxicity, offering a non-amine alternative for pulmonary delivery.
Methodological Strengths
- In vivo demonstration across multiple lung cell types with diverse mRNA cargos.
- Safety assessment indicating no sustained inflammation or tissue damage.
Limitations
- Preclinical murine study; human translatability, dose scaling, and durability remain to be established.
- Long-term immunogenicity and repeated dosing effects were not fully characterized.
Future Directions: Optimization of sLNP chemistry and nebulization for human airways, disease-model testing (e.g., CFTR correction), GLP toxicology, and first-in-human studies for inhaled mRNA therapeutics.
Lung epithelial and immune cells play an important role in respiratory health, serving as the first line of defense. Targeting these cells presents significant therapeutic opportunities, particularly for mRNA-based medicine. However, efficient mRNA delivery to lung cells remains challenging due to mucosal barriers, enzymatic degradation, and complex tissue architecture. In this study, we developed sulfonium lipid nanoparticles (sLNPs) featuring a sulfonium head group and branched tail structure. These sLNPs efficiently delivered mRNA to lung epithelial and immune cells via intranasal instillation in mice, transfecting club cells, ciliated cells, and macrophages, which are key players in lung structure and function. Additionally, sLNPs successfully delivered CRISPR-Cas9 mRNA and sgRNA for genome editing, as well as cytokine mRNA for immune modulation in the lungs. The sLNP platform demonstrated safety in adult mice, with no significant local or systemic tissue damage observed. These findings highlight the sLNP platform's effectiveness and versatility in delivering diverse mRNA molecules, demonstrating its potential for applications ranging from gene editing to immunomodulation therapies. With further optimization, the sLNP system could pave the way for advanced mRNA-based treatments for lung diseases. STATEMENT OF SIGNIFICANCE: Almost all of the previously developed lipids for pulmonary mRNA delivery are amine-based. We designed and synthesized a group of lipids featuring the sulfonium charge-carrying group for mRNA delivery. This is the first demonstration of employing sulfonium lipid nanoparticles (sLNPs) for mRNA delivery to lung epithelial and immune cells in vivo. These sLNPs enabled efficient pulmonary delivery of diverse mRNA cargos, supporting applications such as bioluminescence imaging, gene editing, and immunomodulation. Club and ciliated cells as well as macrophages in the bronchoalveolar fluid, were successfully transfected. No sustained inflammation or toxicity was induced, highlighting the safety of these sulfonium lipid materials.