Daily Respiratory Research Analysis
Three standout studies advance respiratory science and care: a mechanistic study reveals suppressed nasal interferon responses that delay viral clearance and blunt mucosal vaccine immunity; an externally tested deep-learning model markedly improves pulmonary nodule malignancy risk stratification across European screening cohorts; and a new-generation liquid ventilator safely delivers total liquid ventilation in a porcine severe ARDS model, improving survival and gas exchange.
Summary
Three standout studies advance respiratory science and care: a mechanistic study reveals suppressed nasal interferon responses that delay viral clearance and blunt mucosal vaccine immunity; an externally tested deep-learning model markedly improves pulmonary nodule malignancy risk stratification across European screening cohorts; and a new-generation liquid ventilator safely delivers total liquid ventilation in a porcine severe ARDS model, improving survival and gas exchange.
Research Themes
- Mucosal interferon biology and vaccine adjuvant strategies
- AI-driven risk stratification in lung cancer screening
- Innovative ventilation technologies for severe ARDS
Selected Articles
1. Limited nasal IFN production contributes to delayed respiratory virus clearance and suboptimal vaccine responses.
This mechanistic study demonstrates that the nasal mucosa mounts a blunted type I/III interferon response to respiratory viruses, delaying clearance. Early intranasal IFN restored antiviral tone and acted as an effective mucosal vaccine adjuvant, enhancing dendritic cell recruitment, tissue-resident memory T cells, neutralizing antibodies, and reducing clinical disease in HMPV and influenza models.
Impact: It identifies a tractable mechanism for delayed upper airway viral clearance and proposes a readily translatable adjuvant strategy to improve mucosal vaccines.
Clinical Implications: While preclinical, results support testing intranasal type I/III IFN as an early post-exposure intervention and as an adjuvant to strengthen mucosal vaccination against respiratory viruses.
Key Findings
- Nasal airways had higher viral burden with delayed clearance and low type I/III IFN and ISG expression.
- HMPV downregulated nasal IRF3; scRNA-Seq in COVID-19 patients confirmed lower upper-airway ISG signatures.
- Early intranasal type I or III IFN reduced nasal viral titers, increased virus-specific CD8+ T cells, and improved mucosal vaccine responses (enhanced dendritic cells, Trm cells, neutralizing antibodies).
Methodological Strengths
- Integrative design: mouse models, human patient scRNA-Seq validation, and interventional IFN experiments.
- Demonstrated efficacy across two pathogens (HMPV and influenza) and multiple immune readouts.
Limitations
- Predominantly preclinical; no randomized human trials of intranasal IFN adjuvantation.
- Timing and dosing windows for IFN administration may be narrow and require careful optimization.
Future Directions: Conduct early-phase clinical trials of intranasal type I/III IFN for post-exposure prophylaxis and as a mucosal vaccine adjuvant; delineate safety, timing, and dosing and validate in diverse respiratory viruses.
Acute lower respiratory infections are the primary cause of global mortality in postneonatal children. Most respiratory viruses primarily involve upper airway infection and inflammation, yet nasal responses are poorly characterized. Using a mouse model of human metapneumovirus (HMPV), we found viral burden was higher in nasal airways and exhibited delayed clearance. Despite high burden, there was low nasal expression of type I and III interferon (IFN). Single-cell RNA-sequencing (scRNA-Seq) from HMPV-infected mice showed lower nasal IFN-stimulated gene (ISG) expression and nasal enrichment of genes negatively regulating IFN. scRNA-Seq of patients with COVID-19 verified lower ISG expression in upper airways. HMPV infection downregulated nasal expression of IFN regulatory factor 3, suggesting a mechanism for limited response. To rescue the quiescent environment, we administered type I or III IFN to upper airways early postinfection, leading to lower nasal HMPV titer and virus-specific CD8+ T cell upregulation. Intranasal immunization adjuvanted with type I or III IFN improved immune response, reduced clinical disease, and enhanced viral clearance in HMPV and influenza infection. IFN adjuvant increased recruitment of dendritic cells, recruitment of resident memory T cells, and neutralizing antibodies. These findings reveal locally suppressed IFN production contributes to a quiescent nasal immune landscape that delays viral clearance and impairs mucosal vaccine responses.
2. External Test of a Deep Learning Algorithm for Pulmonary Nodule Malignancy Risk Stratification Using European Screening Data.
A deep-learning model trained on NLST generalized well to three European screening trials, achieving AUCs up to 0.98 (1-year cancers) and outperforming or matching the PanCan model across cohorts and subsets of indeterminate nodules. External validation across heterogeneous datasets supports clinical utility in reducing false positives and streamlining management.
Impact: High-quality external validation demonstrates robust generalization of an AI risk model across major screening trials, a critical step toward adoption in lung cancer screening programs.
Clinical Implications: Integrating the DL model into screening workflows could improve malignancy risk stratification for indeterminate nodules, reduce unnecessary follow-up or invasive procedures, and focus resources on high-risk lesions.
Key Findings
- External validation across DLCST, MILD, and NELSON (n=4146; 7614 benign, 180 malignant nodules) showed AUCs 0.98 (1-year), 0.96 (2-year), and 0.94 (overall).
- Performance matched or exceeded the PanCan model, including in indeterminate 5–15 mm nodules and size-matched cancer subsets.
- Generalized across heterogeneous European cohorts despite being trained on U.S. NLST data.
Methodological Strengths
- Large, multicenter external validation with predefined clinically relevant subsets.
- Direct benchmarking against an established clinical model (PanCan).
Limitations
- Retrospective design; potential differences in CT acquisition protocols across trials.
- Limited detail on calibration and decision thresholds for clinical integration.
Future Directions: Prospective impact studies to quantify reduction in false positives, downstream procedures, and cost-effectiveness; calibration and threshold optimization for diverse scanners and populations.
Background Low-dose CT screening reduces lung cancer-related deaths but has high rates of false-positive findings. A deep learning (DL) algorithm could improve nodule risk stratification but requires robust external testing. Purpose To externally test a DL algorithm for nodule malignancy risk estimation using pooled data from three large European lung cancer screening trials. Materials and Methods In this retrospective study, a DL algorithm trained on National Lung Screening Trial data was externally tested using baseline CT scans from the Danish Lung Cancer Screening Trial, the Multicentric Italian Lung Detection trial, and the Dutch-Belgian Lung Cancer Screening Trial. Performance was assessed across the pooled cohort and two subsets: subset A, including indeterminate nodules (5-15 mm); and subset B, including cancers size-matched to benign nodules (1:2 ratio). Performance, including the area under the receiver operating characteristic curve (AUC), was compared with the Pan-Canadian Early Detection of Lung Cancer (PanCan) model. Results The pooled cohort included 4146 participants (median age, 58 years; 78% male participants; median smoking history, 38 pack-years) with 7614 benign and 180 malignant nodules. The DL algorithm achieved AUCs of 0.98, 0.96, and 0.94 for cancers diagnosed within 1 year, 2 years, and throughout screening, respectively, compared with 0.98, 0.94, and 0.93 (
3. Total liquid ventilation in a porcine model of severe acute respiratory distress syndrome using a new generation of liquid ventilator.
Using a new-generation ventilator that precisely controls end-expiratory liquid volume, respiratory rate, and liquid tidal volume, total liquid ventilation was feasible, safe, and improved survival (100% vs 40%) in a swine severe ARDS model. The approach also improved oxygenation and mitigated hypoxemia-related deaths during the experimental period.
Impact: Introduces a clinically relevant large-animal demonstration that controlled TLV can rescue severe ARDS physiology, energizing a long-sought alternative ventilation paradigm.
Clinical Implications: If translated, TLV could provide a rescue ventilation strategy for refractory hypoxemia in ARDS; parameters (EELqV, RR, LqVt) and liquid choice (perfluorocarbon) will be central to safety and efficacy.
Key Findings
- In a severe ARDS swine model, survival was 100% (5/5) with TLV versus 40% (2/5) with continued gas ventilation over 60 minutes.
- The LV4B ventilator maintained tight control of end-expiratory liquid volume, respiratory rate, and liquid tidal volume using perfluorooctyl bromide.
- TLV reduced hypoxemia-related deaths during the experimental window, supporting feasibility and safety in large animals.
Methodological Strengths
- Clinically relevant large-animal (porcine) severe ARDS model with controlled ventilatory parameters.
- Clear survival and physiologic endpoints with an active control arm.
Limitations
- Small sample size (n≈10; 5 per group) and short intervention duration (60 minutes).
- Preclinical study; human safety, dosing schemas, and long-term outcomes remain unknown.
Future Directions: Extend TLV duration and endpoints, assess lung mechanics and injury biomarkers, and progress to early human feasibility trials in refractory ARDS with rigorous safety monitoring.
BACKGROUND: Total liquid ventilation (TLV) has been experimentally proposed as an alternative treatment for the management of Acute Respiratory Distress Syndrome (ARDS). Recent technological advances have led to the evaluation of a TLV prototype in patients resuscitated after cardiac arrest. Here, our goal was to determine whether a derived version of this prototype, so-called LV4B (liquid ventilation for breathing), could be used for normothermic TLV in a swine model of severe ARDS. METHODS: Swine were anesthetized and instrumented for respiratory and hemodynamic evaluation. ARDS was induced by one or two administrations of oleic acid (0.1 mg/kg), until reaching a PaO2/FiO2 ratio < 100 mmHg. After ARDS induction, animals were allocated to undergo 60 min of either gas ventilation continuation (Control group) or TLV using a prototype that continuously controls respiratory rate (RR), liquid tidal volume (LqVt) and end-expiratory liquid volume (EELqV, respectively). Perfluorooctyl bromide was used as breathable liquid. RESULTS: After ARDS induction and group allocation, 2/5 animals (40%) survived in the Control groups versus 5/5 in the TLV group (100%). In the Control group, premature deaths were related to sustained hypoxemia (PaO CONCLUSIONS: TLV with a liquid ventilator controlling EELqV, RR and LqVt is feasible and safe in large animals in a severe model of ARDS. This opens promising perspectives and warrants further investigation, including prolonged treatment durations and long-term follow-up.